Enable unified image budgeting by default (#59282)

## What changed

Mark `unified_image_budget` stable and enable it by default. Supported models now use unified image preparation with `original` detail, and `view_image` omits the `detail` parameter from its advertised schema.

## Testing

Update image input, MCP output, history replay, and tool schema expectations for the new default. Keep explicit coverage of the disabled feature and models that do not support original image detail.

GitOrigin-RevId: 2d6fc8019149fcf18b5813db2d2082ba95de2241
This commit is contained in:
Dmitriy Kovalenko
2026-11-04 01:24:36 +00:00
committed by copyberry
parent 0b21a5b526
commit 0a944f8ce2
105 changed files with 349 additions and 342 deletions
@@ -3022,13 +3022,13 @@ async fn turn_start_reports_personality_overrides_v2(
}
#[tokio::test]
async fn turn_start_defaults_local_image_detail_to_high() -> Result<()> {
async fn turn_start_defaults_local_image_detail_to_original() -> Result<()> {
let input_images = run_local_image_turn(/*detail*/ None).await?;
assert_eq!(input_images.len(), 1);
assert_eq!(
input_images[0].get("detail").and_then(Value::as_str),
Some("high")
Some("original")
);
Ok(())
@@ -6,8 +6,8 @@ Scenario: Guardian follow-up review request layout
## Window 1
-- request 1 (request; Initial Guardian Review Request) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -6,8 +6,8 @@ Scenario: Guardian review request layout
## Window 1
-- request 1 (request; Guardian Review Request) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
+6 -9
View File
@@ -6807,9 +6807,7 @@ image({{
"emit images with legacy detail arguments and MCP metadata",
&code,
"gpt-5.5",
|config| {
let _ = config.features.enable(Feature::UnifiedImageBudget);
},
|_| {},
)
.await?;
@@ -6859,7 +6857,6 @@ async fn code_mode_unified_image_budget_preserves_legacy_contract_for_unsupporte
.with_config(|config| {
let _ = config.features.enable(Feature::CodeMode);
let _ = config.features.enable(Feature::ExecutedToolCallMetadata);
let _ = config.features.enable(Feature::UnifiedImageBudget);
});
let (_test, second_mock) = run_code_mode_turn_with_builder(
&server,
@@ -6946,8 +6943,8 @@ async fn code_mode_can_use_view_image_result_with_image_helper(
.with_model("gpt-5.5")
.with_config(move |config| {
let _ = config.features.enable(Feature::CodeMode);
if unified_image_budget {
let _ = config.features.enable(Feature::UnifiedImageBudget);
if !unified_image_budget {
let _ = config.features.disable(Feature::UnifiedImageBudget);
}
});
let test = builder.build(&server).await?;
@@ -7454,7 +7451,7 @@ async fn code_mode_node_repl_text_evidence_is_visible_only_to_guardian(
let check_detail = enhanced_transcripts && transcript_images && reviewer_constraint.is_none();
let mut large_image = Cursor::new(Vec::new());
if check_detail {
DynamicImage::new_rgba8(/*w*/ 2049, /*h*/ 32)
DynamicImage::new_rgba8(/*w*/ 6001, /*h*/ 32)
.write_to(&mut large_image, ImageFormat::Png)?;
}
let mut builder = test_codex()
@@ -7679,11 +7676,11 @@ await tools.exec_command({ cmd: "printf second", sandbox_permissions: "require_e
.position(|item| item.get("image_url").and_then(Value::as_str) == Some(PRIVATE_IMAGE))
.expect("private reviewer image should be present");
if check_detail {
assert_eq!(reviewer_user_content[image_index]["detail"], "high");
assert_eq!(reviewer_user_content[image_index]["detail"], "original");
let payload = reviewer_image_urls[1].split_once(',').unwrap().1;
let dimensions =
image::load_from_memory(&BASE64_STANDARD.decode(payload)?)?.dimensions();
assert_eq!(dimensions, (2048, 32));
assert_eq!(dimensions, (6000, 32));
}
for (index, marker) in [(image_index - 1, "before"), (image_index + 1, "after")] {
let text = reviewer_user_content[index]["text"].as_str().unwrap();
+8 -7
View File
@@ -265,7 +265,7 @@ async fn copy_paste_local_image_persists_rollout_request_shape() -> anyhow::Resu
},
ContentItem::InputImage {
image: ImageReference::Inline { image_url },
detail: Some(DEFAULT_IMAGE_DETAIL),
detail: Some(ImageDetail::Original),
},
ContentItem::InputText {
text: codex_protocol::models::image_close_tag_text(),
@@ -360,7 +360,7 @@ async fn drag_drop_image_persists_rollout_request_shape() -> anyhow::Result<()>
content: vec![
ContentItem::InputImage {
image: ImageReference::Inline { image_url },
detail: Some(DEFAULT_IMAGE_DETAIL),
detail: Some(ImageDetail::Original),
},
ContentItem::InputText {
text: "dropped image".to_string(),
@@ -640,8 +640,8 @@ async fn resumed_history_only_emits_resize_notices_for_new_images() -> anyhow::R
let image_path = initial.cwd.path().join("large-image.png");
ImageBuffer::from_pixel(
/*width*/ 2304,
/*height*/ 864,
/*width*/ 6400,
/*height*/ 256,
Rgba([12u8, 34, 56, 255]),
)
.save(&image_path)?;
@@ -728,7 +728,7 @@ async fn resumed_history_only_emits_resize_notices_for_new_images() -> anyhow::R
);
assert_eq!(
image::load_from_memory(&upload.data)?.dimensions(),
(2048, 768)
(6000, 240)
);
}
@@ -769,7 +769,8 @@ async fn resumed_history_only_emits_resize_notices_for_new_images() -> anyhow::R
.split_once(',')
.context("historical image data URL")?;
let historical_image = image::load_from_memory(&BASE64_STANDARD.decode(encoded_image)?)?;
assert_eq!(historical_image.dimensions(), (2048, 768));
// Replay honors the high detail recorded with the historical image.
assert_eq!(historical_image.dimensions(), (2048, 82));
assert_eq!(
input[image_message_indices[1]]
.get("content")
@@ -782,7 +783,7 @@ async fn resumed_history_only_emits_resize_notices_for_new_images() -> anyhow::R
let expected_notice = concat!(
"<image_resize_notice>\n",
"Image 1 of 1 in the preceding user message was resized from 2304x864 to 2048x768 pixels.\n",
"Image 1 of 1 in the preceding user message was resized from 6400x256 to 6000x240 pixels.\n",
"</image_resize_notice>"
);
let resize_notices = request
+3 -3
View File
@@ -3435,7 +3435,7 @@ async fn stdio_image_responses_resize_large_image() -> anyhow::Result<()> {
let server_name = "rmcp";
let namespace = format!("mcp__{server_name}");
let original_dimensions = (3000, 2000);
let original_dimensions = (6400, 320);
let image = ImageBuffer::from_pixel(
original_dimensions.0,
original_dimensions.1,
@@ -3509,14 +3509,14 @@ async fn stdio_image_responses_resize_large_image() -> anyhow::Result<()> {
let resized_url = output[1]["image_url"]
.as_str()
.expect("MCP image output should contain a data URL");
assert_eq!(output[1]["detail"], "high");
assert_eq!(output[1]["detail"], "original");
let (_, resized_base64) = resized_url
.split_once(',')
.expect("resized image should contain a data URL prefix");
let resized_bytes = BASE64_STANDARD.decode(resized_base64)?;
let resized = image::load_from_memory(&resized_bytes)?;
let resized_dimensions = resized.dimensions();
assert_eq!(resized_dimensions, (1920, 1280));
assert_eq!(resized_dimensions, (6000, 300));
server.verify().await;
Ok(())
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (11; hash=62414ae13aabdb67):
tools (11; hash=147c1017bfad1ffd):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/multi_agent_v1: Tools for spawning and managing sub-agents.; hash=2327142bda177c50
- function/close_agent
- function/resume_agent
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (11; hash=62414ae13aabdb67):
tools (11; hash=147c1017bfad1ffd):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/multi_agent_v1: Tools for spawning and managing sub-agents.; hash=2327142bda177c50
- function/close_agent
- function/resume_agent
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (11; hash=62414ae13aabdb67):
tools (11; hash=147c1017bfad1ffd):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/multi_agent_v1: Tools for spawning and managing sub-agents.; hash=2327142bda177c50
- function/close_agent
- function/resume_agent
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (11; hash=62414ae13aabdb67):
tools (11; hash=147c1017bfad1ffd):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/multi_agent_v1: Tools for spawning and managing sub-agents.; hash=2327142bda177c50
- function/close_agent
- function/resume_agent
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (11; hash=62414ae13aabdb67):
tools (11; hash=147c1017bfad1ffd):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/multi_agent_v1: Tools for spawning and managing sub-agents.; hash=2327142bda177c50
- function/close_agent
- function/resume_agent
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (11; hash=62414ae13aabdb67):
tools (11; hash=147c1017bfad1ffd):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/multi_agent_v1: Tools for spawning and managing sub-agents.; hash=2327142bda177c50
- function/close_agent
- function/resume_agent
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (11; hash=62414ae13aabdb67):
tools (11; hash=147c1017bfad1ffd):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/multi_agent_v1: Tools for spawning and managing sub-agents.; hash=2327142bda177c50
- function/close_agent
- function/resume_agent
@@ -7,8 +7,8 @@ Scenario: A parent with no configured environment delegates an API migration, qu
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (2; hash=7958a01e34d95ac2):
- namespace/functions; hash=30a5fc1fcffb1c04
00:additional_tools/developer (2; hash=736be7a0054c7ff6):
- namespace/functions; hash=f33f647fee88e25e
- custom/exec
- function/wait
- function/request_user_input
@@ -66,8 +66,8 @@ Scenario: A parent with no configured environment delegates an API migration, qu
## Window 2 (after request 6: input diverged at item 03)
-- request 7 (turn) --
00:additional_tools/developer (2; hash=7958a01e34d95ac2):
- namespace/functions; hash=30a5fc1fcffb1c04
00:additional_tools/developer (2; hash=736be7a0054c7ff6):
- namespace/functions; hash=f33f647fee88e25e
- custom/exec
- function/wait
- function/request_user_input
@@ -6,8 +6,8 @@ Scenario: Permission discovery preserves legacy output, exposes reply-only acces
## Window 1
-- request 1 (request) --
00:additional_tools/developer (3; hash=788c4a5014dcfe4f):
- namespace/functions; hash=c8b8fa91706a7620
00:additional_tools/developer (3; hash=ffbbdab1ef8f8687):
- namespace/functions; hash=5fe7992d16964702
- function/exec_command
- function/write_stdin
- function/request_user_input
@@ -15,8 +15,8 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (4; hash=aa7c4090dac68d7b):
- custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=9c848ebf633880a2
tools (4; hash=1f8fea8211b3c63d):
- custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=7f56c3b386ed268e
- function/wait: Waits on a yielded `exec` cell and returns new output or completion. - Use `wait` only after `exe...; args=[cell_id, max_tokens, terminate, yield_time_ms]; hash=6692d8f6c55be489
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- web_search; hash=e214d4db1bc49e89
@@ -6,8 +6,8 @@ Scenario: Astra starts in one workspace, switches for the next step of the activ
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (3; hash=62160ed28e692c35):
- namespace/functions; hash=4261ffb241688b3d
00:additional_tools/developer (3; hash=73b17e628bb8aee1):
- namespace/functions; hash=7c1b7203dcaa5b6f
- custom/exec
- function/wait
- function/request_user_input
@@ -6,8 +6,8 @@ Scenario: Astra compacts a completed launch-notes turn in the background. The us
## Window 1
-- request 1 (request) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -45,8 +45,8 @@ Scenario: Astra compacts a completed launch-notes turn in the background. The us
## Window 2 (after request 2: input diverged at item 10)
-- request 3 (request) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -88,8 +88,8 @@ Scenario: Astra compacts a completed launch-notes turn in the background. The us
## Window 3 (after request 3: input diverged at item 02)
-- request 4 (request) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -6,8 +6,8 @@ Scenario: Astra uses the catalog question schema to ask who a launch update is f
## Window 1
-- request 1 (request) --
00:additional_tools/developer (3; hash=f181967456befa67):
- namespace/functions; hash=7e611b8ddd5621cf
00:additional_tools/developer (3; hash=42141f5d3211bed3):
- namespace/functions; hash=80c4fbeb0cbc76f5
- custom/exec
- function/wait
- function/request_user_input
@@ -6,8 +6,8 @@ Scenario: Astra reads code-mode timings and the collector notice even at zero to
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (request) --
00:additional_tools/developer (3; hash=5e8370540461db70):
- namespace/functions; hash=7d26e086c1dbba28
00:additional_tools/developer (3; hash=90063690a488dd0c):
- namespace/functions; hash=596780ea93468a82
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (request) --
00:additional_tools/developer (3; hash=5e8370540461db70):
- namespace/functions; hash=7d26e086c1dbba28
00:additional_tools/developer (3; hash=90063690a488dd0c):
- namespace/functions; hash=596780ea93468a82
- custom/exec
- function/wait
- function/request_user_input
@@ -6,8 +6,8 @@ Scenario: Astra receives new user input after its unfinished response is interru
## Window 1
-- request 1 (request) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -6,8 +6,8 @@ Scenario: Astra continues after new user input yields a running code cell, then
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (3; hash=69c8f596545517aa):
- namespace/functions; hash=13a36c9bc9ada856
00:additional_tools/developer (3; hash=77f06b6af2d274e6):
- namespace/functions; hash=9babd62cd00098f4
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -188,8 +188,8 @@ Settings:
## Window 2 (after request 3: input diverged at item 02)
Settings: same as window 1
-- request 4 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -6,8 +6,8 @@ Scenario: A host contributes budget reminders at thresholds. Remote compaction s
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -54,8 +54,8 @@ Scenario: A host contributes budget reminders at thresholds. Remote compaction s
## Window 2 (after request 4: input diverged at item 02)
-- request 5 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -15,12 +15,12 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (7; hash=994c9baf5d9a36a0):
tools (7; hash=4bc85f667bb09668):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- tool_search: # Tool discovery Searches over deferred tool metadata with BM25 and exposes matching tools for t...; args=[limit, query]; hash=b573ed69c0b0e6c7
- web_search; hash=f103e327c6870a46
-- request 1 (turn; Initial bundled guidance) --
@@ -88,7 +88,7 @@ Settings:
## Window 2 (after request 1)
Settings: relative to window 1
model: "plugins-a"
tools (10; hash=df5c5997ff843305):
tools (10; hash=9f8074d045f3bd93):
+ function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
+ function/list_mcp_resource_templates: Lists resource templates provided by MCP servers. Parameterized resource templates allow servers ...; args=[cursor, server]; hash=60f17c90ef7a4ebb
+ function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (10; hash=df5c5997ff843305):
tools (10; hash=9f8074d045f3bd93):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- tool_search: # Tool discovery Searches over deferred tool metadata with BM25 and exposes matching tools for t...; args=[limit, query]; hash=b573ed69c0b0e6c7
- web_search; hash=f103e327c6870a46
-- request 1 (turn) --
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (11; hash=70f020489861129c):
tools (11; hash=2123887d18ce7e32):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/multi_agent_v1: Tools for spawning and managing sub-agents.; hash=2327142bda177c50
- function/close_agent
- function/resume_agent
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (12; hash=222ce6d65d94fe37):
tools (12; hash=2a0c7daad939067d):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/multi_agent_v1: Tools for spawning and managing sub-agents.; hash=2327142bda177c50
- function/close_agent
- function/resume_agent
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (12; hash=222ce6d65d94fe37):
tools (12; hash=2a0c7daad939067d):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/multi_agent_v1: Tools for spawning and managing sub-agents.; hash=2327142bda177c50
- function/close_agent
- function/resume_agent
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=37af7b347b5c1eb7):
- namespace/functions; hash=a3c60999d0288751
00:additional_tools/developer (3; hash=72a6498aa4648fdf):
- namespace/functions; hash=6fcc6902881b24af
- custom/exec
- function/wait
- function/request_user_input
@@ -54,8 +54,8 @@ Settings:
## Window 2 (after request 3: input diverged at item 00)
Settings: same as window 1
-- request 4 (turn) --
00:additional_tools/developer (3; hash=ce2de39d992cd930):
- namespace/functions; hash=fa0731964313a476
00:additional_tools/developer (3; hash=cc5960827562f0c8):
- namespace/functions; hash=9865166c733563f0
- custom/exec
- function/wait
- function/request_user_input
@@ -15,8 +15,8 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (7; hash=7916dd640b46ec6e):
- custom/exec: Catalog exec instructions. Default yield: 1234 ms. `image(imageUrlOrItem: string | { image_url: s...; hash=a49b91fd82908e18
tools (7; hash=3eab8d5a4c823294):
- custom/exec: Catalog exec instructions. Default yield: 1234 ms. `image(imageUrlOrItem: string | { image_url: s...; hash=2e2d9d0d2c0430f0
- function/wait: Catalog wait instructions: resume the yielded cell.; args=[cell_id, max_tokens, terminate, yield_time_ms]; hash=67bc0f04da3505b3
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- function/request_user_input_async: Ask the user one or more questions during ongoing work. Use this tool only to request missing inf...; args=[questions]; hash=d6d74f33db330162
@@ -15,8 +15,8 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (7; hash=40e072f613a5e3eb):
- custom/exec: Catalog exec instructions. Default yield: 1234 ms. `image(imageUrlOrItem: string | { image_url: s...; hash=1b186fc5b7e14af3
tools (7; hash=833e88d9ec8967f7):
- custom/exec: Catalog exec instructions. Default yield: 1234 ms. `image(imageUrlOrItem: string | { image_url: s...; hash=42b537270d73a5d1
- function/wait: Catalog wait instructions: resume the yielded cell.; args=[cell_id, max_tokens, terminate, yield_time_ms]; hash=67bc0f04da3505b3
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- function/request_user_input_async: Ask the user one or more questions during ongoing work. Use this tool only to request missing inf...; args=[questions]; hash=d6d74f33db330162
@@ -15,8 +15,8 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (4; hash=95b65d00fe586b7d):
- custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=e02c9856c9cd8430
tools (4; hash=5e7a9f3f74e9553f):
- custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=f38c8248ec696414
- function/wait: Waits on a yielded `exec` cell and returns new output or completion. - Use `wait` only after `exe...; args=[cell_id, max_tokens, terminate, yield_time_ms]; hash=6692d8f6c55be489
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- web_search; hash=e214d4db1bc49e89
@@ -6,8 +6,8 @@ Scenario: Settlement helpers deliver fulfilled and rejected records with input i
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn; Before /compact) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -109,8 +109,8 @@ Settings:
## Window 2 (after request 2: input diverged at item 07)
Settings: unavailable (items only)
-- request 3 (items; Stored replacement before next turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -198,8 +198,8 @@ Settings: unavailable (items only)
## Window 3 (after request 3)
Settings: same as window 1
-- request 4 (turn; First turn after /compact) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -15,14 +15,14 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (9; hash=351b20e3aa629305):
tools (9; hash=5edab7fdbfe2dd93):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- function/new_context: Catalog context guidance.; args=[]; hash=8d300f346fb11d84
- function/get_context_remaining: Catalog context guidance.; args=[]; hash=802b1cf4a796b439
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- tool_search: # Tool discovery Searches over deferred tool metadata with BM25 and exposes matching tools for t...; args=[limit, query]; hash=b573ed69c0b0e6c7
- web_search; hash=f103e327c6870a46
-- request 1 (turn) --
@@ -50,7 +50,7 @@ Settings:
## Window 2 (after request 1)
Settings: relative to window 1
model: "context-tools-updated"
tools (9; hash=6f6db580112d2227):
tools (9; hash=829060c0abfc8021):
~ function/new_context; args=[]; hash=b9263827b0adbb78
~ function/get_context_remaining; args=[]; hash=3e96d66806c56bdd
-- request 2 (turn) --
@@ -16,7 +16,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (10; hash=39aa00a8736bb135):
tools (10; hash=b0582dd610604bdb):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
@@ -25,7 +25,7 @@ Settings:
- function/curr_time
- function/sleep
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/collaboration: Tools for spawning and managing sub-agents.; hash=2c8d6295978dbe63
- function/followup_task
- function/interrupt_agent
@@ -51,7 +51,7 @@ Settings:
## Window 2 (after request 1)
Settings: relative to window 1
tools (10; hash=f9e31c2361b9bc7d):
tools (10; hash=d346da81584ec2d3):
- removed function/release
+ function/deployment: Check deployment.; args=[]; hash=2e9a71e9a20e43ad
-- request 2 (turn) --
@@ -81,7 +81,7 @@ Settings: relative to window 1
## Window 3 (after request 4)
Settings: relative to window 2
tools (9; hash=4f2e62909d4cf2eb):
tools (9; hash=79c1c8bb10577e35):
- removed function/deployment
-- request 5 (turn) --
00:message/developer:
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (request) --
00:additional_tools/developer (3; hash=e075a43448be3246):
- namespace/functions; hash=bbcd688349ca7e12
00:additional_tools/developer (3; hash=c989fc4f95fb6f22):
- namespace/functions; hash=80f0e6670cf933cc
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=650d9374d11dab9f):
- namespace/functions; hash=101d32aa59ea258b
00:additional_tools/developer (3; hash=cec010d63fda60d5):
- namespace/functions; hash=106dd1b983efd2f7
- function/exec_command
- function/write_stdin
- function/request_user_input
@@ -25,11 +25,15 @@ Settings:
- function/view_image
- namespace/image_gen: Tools in the image_gen namespace.; hash=7f9a214246396279
- function/imagegen
- tool_search: # Tool discovery Searches over deferred tool metadata with BM25 and exposes matching tools for t...; args=[limit, query]; hash=400cd6ac61d524a9
- tool_search: # Tool discovery Searches over deferred tool metadata with BM25 and exposes matching tools for t...; args=[limit, query]; hash=b573ed69c0b0e6c7
01:message/developer:
<MODEL_INSTRUCTIONS>
02:message/developer:
<PERMISSIONS_INSTRUCTIONS>
02:message/developer[2]:
[01] <PERMISSIONS_INSTRUCTIONS>
[02] <tools>
Deferred tool namespaces:
- multi_agent_v1: Tools for spawning and managing sub-agents.
</tools>
03:message/user:
<ENVIRONMENT_CONTEXT>
04:message/user:
@@ -80,8 +80,8 @@ Scenario: A historical MAv2 send_message call and an encrypted parent reply reac
## Window 3 (after request 3: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 4 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -200,8 +200,8 @@ Scenario: A historical MAv2 send_message call and an encrypted parent reply reac
## Window 5 (after request 5: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 6 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -80,8 +80,8 @@ Scenario: A historical MAv2 send_message call and an encrypted parent reply reac
## Window 3 (after request 3: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 4 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -200,8 +200,8 @@ Scenario: A historical MAv2 send_message call and an encrypted parent reply reac
## Window 5 (after request 5: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 6 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -41,8 +41,8 @@ Scenario: An encrypted parent reply is delivered as a native agent message. Both
## Window 2 (after request 2: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 3 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -149,8 +149,8 @@ Scenario: An encrypted parent reply is delivered as a native agent message. Both
## Window 4 (after request 4: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 5 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -15,12 +15,12 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (7; hash=f47d2e65b8d3bd06):
tools (7; hash=c8fad8866d479502):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=7c8f464569dbaa9c
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- tool_search: # Tool discovery Searches over deferred tool metadata with BM25 and exposes matching tools for t...; args=[limit, query]; hash=b573ed69c0b0e6c7
- web_search; hash=f103e327c6870a46
-- request 1 (turn) --
@@ -54,8 +54,8 @@ Settings: relative to window 1
reasoning: {"context":"all_turns","effort":"medium"}
tools: <removed>
-- request 4 (turn) --
00:additional_tools/developer (1; hash=1ea3c73f7e3c8b33):
- namespace/functions; hash=2a9848b19a516663
00:additional_tools/developer (1; hash=4c5eea94c78f0b7f):
- namespace/functions; hash=13ba6f301e19dc45
- custom/exec
- function/wait
- function/request_user_input
@@ -87,8 +87,8 @@ Settings: relative to window 2
text: {"format":{"name":"codex_output_schema","schema":{"additionalProperties":false,"properties":{"outcome":{"enum":["allow","deny"],"type":"string"},"rationale":{"type":"string"},"risk_level":{"enum":["low","medium","high","critical"],"type":"string"},"user_authorization":{"enum":["unknown","low","medium","high"],"type":"string"}},"required":["outcome"],"type":"object"},"strict":false,"type":"json_schema"},"verbosity":"low"}
-- request 6 (turn) --
00:compaction:encrypted=true; chars=39; hash=4010be12538dbedf
01:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
01:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
02:message/developer:
@@ -185,8 +185,8 @@ Settings: relative to window 2
## Window 4 (after request 6: input diverged at item 00)
Settings: same as window 2
-- request 7 (turn) --
00:additional_tools/developer (1; hash=1ea3c73f7e3c8b33):
- namespace/functions; hash=2a9848b19a516663
00:additional_tools/developer (1; hash=4c5eea94c78f0b7f):
- namespace/functions; hash=13ba6f301e19dc45
- custom/exec
- function/wait
- function/request_user_input
@@ -217,8 +217,8 @@ Settings: same as window 2
## Window 5 (after request 8: input diverged at item 03)
Settings: same as window 2
-- request 9 (turn) --
00:additional_tools/developer (1; hash=1ea3c73f7e3c8b33):
- namespace/functions; hash=2a9848b19a516663
00:additional_tools/developer (1; hash=4c5eea94c78f0b7f):
- namespace/functions; hash=13ba6f301e19dc45
- custom/exec
- function/wait
- function/request_user_input
@@ -248,8 +248,8 @@ Settings: same as window 2
Settings: same as window 3
-- request 10 (turn) --
00:compaction:encrypted=true; chars=33; hash=5c07b4383c427607
01:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
01:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
02:message/developer:
@@ -345,8 +345,8 @@ Settings: same as window 3
## Window 7 (after request 10: input diverged at item 00)
Settings: same as window 2
-- request 11 (turn) --
00:additional_tools/developer (1; hash=1ea3c73f7e3c8b33):
- namespace/functions; hash=2a9848b19a516663
00:additional_tools/developer (1; hash=4c5eea94c78f0b7f):
- namespace/functions; hash=13ba6f301e19dc45
- custom/exec
- function/wait
- function/request_user_input
@@ -118,8 +118,8 @@ Scenario: A root uses Code Mode to send a confirmation and receives a real user
## Window 3 (after request 6: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 7 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -188,8 +188,8 @@ Scenario: A root uses Code Mode to send a confirmation and receives a real user
## Window 4 (after request 7: input diverged at item 04, settings changed (prompt_cache_key))
-- request 8 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -7,8 +7,8 @@ Scenario: Direct Guardian review of an injected exec call; an allow assessment d
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -21,8 +21,8 @@ Scenario: Live and direct Guardian review of a patch; direct approval leaves the
## Window 2 (after request 1: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 2 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -103,8 +103,8 @@ Scenario: Live and direct Guardian review of a patch; direct approval leaves the
## Window 4 (after request 3: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 4 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -21,8 +21,8 @@ Scenario: Live and direct Guardian review of a patch; direct approval leaves the
## Window 2 (after request 1: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 2 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -113,8 +113,8 @@ Scenario: Live and direct Guardian review of a patch; direct approval leaves the
## Window 4 (after request 3: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 4 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -22,8 +22,8 @@ Scenario: Live and direct Guardian review receive the same permission request.
## Window 2 (after request 1: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 2 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -107,8 +107,8 @@ Scenario: Live and direct Guardian review receive the same permission request.
## Window 4 (after request 3: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 4 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -6,8 +6,8 @@ Scenario: Direct Guardian review applies the PreToolUse rewrite before checking
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -15,12 +15,12 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (7; hash=994c9baf5d9a36a0):
tools (7; hash=4bc85f667bb09668):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- tool_search: # Tool discovery Searches over deferred tool metadata with BM25 and exposes matching tools for t...; args=[limit, query]; hash=b573ed69c0b0e6c7
- web_search; hash=f103e327c6870a46
-- request 1 (turn) --
@@ -88,8 +88,8 @@ Settings: relative to window 1
text: {"format":{"name":"codex_output_schema","schema":{"additionalProperties":false,"properties":{"outcome":{"enum":["allow","deny"],"type":"string"},"rationale":{"type":"string"},"risk_level":{"enum":["low","medium","high","critical"],"type":"string"},"user_authorization":{"enum":["unknown","low","medium","high"],"type":"string"}},"required":["outcome"],"type":"object"},"strict":false,"type":"json_schema"},"verbosity":"low"}
tools: <removed>
-- request 2 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -7,8 +7,8 @@ Scenario: Complete Guardian request history: root delegates through alpha to lea
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -7,8 +7,8 @@ Scenario: Complete Guardian request history: root delegates through alpha to lea
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -62,8 +62,8 @@ Scenario: Three scheduled runs share one instruction body, a human overrides its
## Window 2 (after request 5: input diverged at item 00, settings changed (model, parallel_tool_calls, prompt_cache_key, reasoning, text, tools))
-- request 6 (request) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -15,12 +15,12 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (7; hash=994c9baf5d9a36a0):
tools (7; hash=4bc85f667bb09668):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- tool_search: # Tool discovery Searches over deferred tool metadata with BM25 and exposes matching tools for t...; args=[limit, query]; hash=b573ed69c0b0e6c7
- web_search; hash=f103e327c6870a46
-- request 1 (turn) --
@@ -46,8 +46,8 @@ Settings: relative to window 1
text: {"format":{"name":"codex_output_schema","schema":{"additionalProperties":false,"properties":{"outcome":{"enum":["allow","deny"],"type":"string"},"rationale":{"type":"string"},"risk_level":{"enum":["low","medium","high","critical"],"type":"string"},"user_authorization":{"enum":["unknown","low","medium","high"],"type":"string"}},"required":["outcome"],"type":"object"},"strict":false,"type":"json_schema"},"verbosity":"low"}
tools: <removed>
-- request 2 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -15,12 +15,12 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (7; hash=994c9baf5d9a36a0):
tools (7; hash=4bc85f667bb09668):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- tool_search: # Tool discovery Searches over deferred tool metadata with BM25 and exposes matching tools for t...; args=[limit, query]; hash=b573ed69c0b0e6c7
- web_search; hash=f103e327c6870a46
-- request 1 (turn) --
@@ -92,8 +92,8 @@ Settings: relative to window 3
text: {"format":{"name":"codex_output_schema","schema":{"additionalProperties":false,"properties":{"outcome":{"enum":["allow","deny"],"type":"string"},"rationale":{"type":"string"},"risk_level":{"enum":["low","medium","high","critical"],"type":"string"},"user_authorization":{"enum":["unknown","low","medium","high"],"type":"string"}},"required":["outcome"],"type":"object"},"strict":false,"type":"json_schema"},"verbosity":"low"}
tools: <removed>
-- request 6 (turn) --
00:additional_tools/developer (1; hash=65a86311792f6c71):
- namespace/functions; hash=dcfbcf0da1c2d607
00:additional_tools/developer (1; hash=7758f662cd124827):
- namespace/functions; hash=70d14ec929ba5cd7
- custom/exec
- function/wait
01:message/developer:
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=7748a53142f449b7):
- namespace/functions; hash=124fee69a69ed6b2
00:additional_tools/developer (3; hash=c6476125e1d67bb3):
- namespace/functions; hash=6475578c7f948d18
- function/exec_command
- function/write_stdin
- function/list_mcp_resources
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (2; hash=b07492b5514a41d0):
- namespace/functions; hash=cc6e8cf0e5405f3a
00:additional_tools/developer (2; hash=019c865b71a05fec):
- namespace/functions; hash=f3355f5705e28592
- custom/exec
- function/wait
- function/exec_command
@@ -46,8 +46,8 @@ Settings:
Settings: relative to window 1
model: "updated-tools-model"
-- request 3 (turn) --
00:additional_tools/developer (2; hash=b07492b5514a41d0):
- namespace/functions; hash=cc6e8cf0e5405f3a
00:additional_tools/developer (2; hash=019c865b71a05fec):
- namespace/functions; hash=f3355f5705e28592
- custom/exec
- function/wait
- function/exec_command
@@ -93,8 +93,8 @@ Settings: relative to window 1
## Window 3 (after request 5: input diverged at item 00)
Settings: same as window 2
-- request 6 (turn) --
00:additional_tools/developer (2; hash=f9f0840b2e42f7f5):
- namespace/functions; hash=9d5d7a0adab05fef
00:additional_tools/developer (2; hash=7df544e43dd38360):
- namespace/functions; hash=1e51a9a9273be2f8
- custom/exec
- function/wait
- function/exec_command
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=7748a53142f449b7):
- namespace/functions; hash=124fee69a69ed6b2
00:additional_tools/developer (3; hash=c6476125e1d67bb3):
- namespace/functions; hash=6475578c7f948d18
- function/exec_command
- function/write_stdin
- function/list_mcp_resources
@@ -15,14 +15,14 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (9; hash=7ebf49dda2901fbc):
- custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=3b60611486072e76
tools (9; hash=c36c84db667eb4c8):
- custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=9226aa5c0caf3b05
- function/wait: Waits on a yielded `exec` cell and returns new output or completion. - Use `wait` only after `exe...; args=[cell_id, max_tokens, terminate, yield_time_ms]; hash=6692d8f6c55be489
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction. exec tool dec...; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=759843fc2a7d41a0
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output. exec tool decla...; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=e6405481b30a43c1
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=11c5b5ed4fb972e8
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=1b87dd8874f6fa10
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=0e7abbd0a5b17fb9
- tool_search: # Tool discovery Searches over deferred tool metadata with BM25 and exposes matching tools for t...; args=[limit, query]; hash=b573ed69c0b0e6c7
- web_search; hash=e214d4db1bc49e89
-- request 1 (turn) --
@@ -45,7 +45,7 @@ Settings:
## Window 2 (after request 2)
Settings: relative to window 1
model: "updated-tools-model"
tools (9; hash=6531bfd30ee7d343):
tools (9; hash=adb28e07b75745e2):
~ custom/exec: Updated execution instructions. Some deferred nested tools may be omitted from this description....; hash=7af6961848c10849
-- request 3 (turn) --
00:message/developer:
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (2; hash=b07492b5514a41d0):
- namespace/functions; hash=cc6e8cf0e5405f3a
00:additional_tools/developer (2; hash=019c865b71a05fec):
- namespace/functions; hash=f3355f5705e28592
- custom/exec
- function/wait
- function/exec_command
@@ -46,8 +46,8 @@ Settings:
Settings: relative to window 1
model: "updated-tools-model"
-- request 3 (turn) --
00:additional_tools/developer (2; hash=b07492b5514a41d0):
- namespace/functions; hash=cc6e8cf0e5405f3a
00:additional_tools/developer (2; hash=019c865b71a05fec):
- namespace/functions; hash=f3355f5705e28592
- custom/exec
- function/wait
- function/exec_command
@@ -92,8 +92,8 @@ Settings: relative to window 1
## Window 3 (after request 5: input diverged at item 00)
Settings: same as window 2
-- request 6 (turn) --
00:additional_tools/developer (2; hash=f9f0840b2e42f7f5):
- namespace/functions; hash=9d5d7a0adab05fef
00:additional_tools/developer (2; hash=7df544e43dd38360):
- namespace/functions; hash=1e51a9a9273be2f8
- custom/exec
- function/wait
- function/exec_command
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (1; hash=a523a0cf594329a1):
- namespace/functions; hash=d3e1dde6718dc7a3
00:additional_tools/developer (1; hash=638dc23247f41f35):
- namespace/functions; hash=f730aa90585570c5
- custom/exec
- function/wait
- function/request_user_input
@@ -40,8 +40,8 @@ Settings:
## Window 2 (after request 2: input diverged at item 05)
Settings: same as window 1
-- request 3 (turn) --
00:additional_tools/developer (1; hash=a523a0cf594329a1):
- namespace/functions; hash=d3e1dde6718dc7a3
00:additional_tools/developer (1; hash=638dc23247f41f35):
- namespace/functions; hash=f730aa90585570c5
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (2; hash=7b4c6a992fd91f91):
- namespace/functions; hash=101d32aa59ea258b
00:additional_tools/developer (2; hash=ba36cb129fbde5df):
- namespace/functions; hash=106dd1b983efd2f7
- function/exec_command
- function/write_stdin
- function/request_user_input
@@ -56,8 +56,8 @@ Settings:
## Window 2 (after request 3: input diverged at item 00)
Settings: same as window 1
-- request 4 (turn) --
00:additional_tools/developer (2; hash=3b0da5c6c5e4fd8e):
- namespace/functions; hash=36c9912d971c5be8
00:additional_tools/developer (2; hash=4f95a0a93c66dfca):
- namespace/functions; hash=e61b39bab731172a
- function/update_plan
- function/request_user_input
- custom/apply_patch
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (2; hash=b07492b5514a41d0):
- namespace/functions; hash=cc6e8cf0e5405f3a
00:additional_tools/developer (2; hash=019c865b71a05fec):
- namespace/functions; hash=f3355f5705e28592
- custom/exec
- function/wait
- function/exec_command
@@ -49,8 +49,8 @@ Settings:
## Window 2 (after request 3: input diverged at item 02)
Settings: same as window 1
-- request 4 (turn) --
00:additional_tools/developer (2; hash=b07492b5514a41d0):
- namespace/functions; hash=cc6e8cf0e5405f3a
00:additional_tools/developer (2; hash=019c865b71a05fec):
- namespace/functions; hash=f3355f5705e28592
- custom/exec
- function/wait
- function/exec_command
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (2; hash=f1d0adb3566d1b59):
- namespace/functions; hash=dfb4b8684addc20d
00:additional_tools/developer (2; hash=64bcbde2bd6a71cd):
- namespace/functions; hash=bf2bbed4b1e4d23d
- custom/exec
- function/wait
- function/exec_command
@@ -57,8 +57,8 @@ Settings:
## Window 2 (after request 2: input diverged at item 03)
Settings: same as window 1
-- request 3 (turn) --
00:additional_tools/developer (2; hash=f1d0adb3566d1b59):
- namespace/functions; hash=dfb4b8684addc20d
00:additional_tools/developer (2; hash=64bcbde2bd6a71cd):
- namespace/functions; hash=bf2bbed4b1e4d23d
- custom/exec
- function/wait
- function/exec_command
@@ -15,8 +15,8 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (4; hash=1261f9f79ad97cee):
- custom/exec: Exec guidance. Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided Ja...; hash=db96605f80dbbdd7
tools (4; hash=1873c769e1ecc1e6):
- custom/exec: Exec guidance. Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided Ja...; hash=4b18b0a399fb90e1
- function/wait: Waits on a yielded `exec` cell and returns new output or completion. - Use `wait` only after `exe...; args=[cell_id, max_tokens, terminate, yield_time_ms]; hash=6692d8f6c55be489
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- web_search; hash=e214d4db1bc49e89
@@ -44,13 +44,13 @@ Settings:
07:custom_tool_call_output:Script completed
Wall time <DURATION> seconds
Output:
| {"descriptions":["Functions context.\\n\\nImage guidance.\\n\\nView a local image file from the filesystem w...>>; };\\n```"],"echo":"ECHOING: ping","agent":"/root"} [hash=4bd42f2e71ce15d7]
| {"descriptions":["Functions context.\\n\\nImage guidance.\\n\\nView a local image file from the filesystem w...>>; };\\n```"],"echo":"ECHOING: ping","agent":"/root"} [hash=e3b76c8d361418c6]
## Window 2 (after request 2)
Settings: relative to window 1
model: "prefixes-cleared"
tools (4; hash=d878b195101dccae):
~ custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=881b443f5068d3fb
tools (4; hash=7831e6ce06adc804):
~ custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=b2642100ba673aff
-- request 3 (turn) --
00:message/developer:
<MODEL_INSTRUCTIONS>
@@ -74,7 +74,7 @@ Settings: relative to window 1
07:custom_tool_call_output:Script completed
Wall time <DURATION> seconds
Output:
| {"descriptions":["Functions context.\\n\\nImage guidance.\\n\\nView a local image file from the filesystem w...>>; };\\n```"],"echo":"ECHOING: ping","agent":"/root"} [hash=4bd42f2e71ce15d7]
| {"descriptions":["Functions context.\\n\\nImage guidance.\\n\\nView a local image file from the filesystem w...>>; };\\n```"],"echo":"ECHOING: ping","agent":"/root"} [hash=e3b76c8d361418c6]
08:message/assistant:
Done.
09:message/developer:
@@ -92,4 +92,4 @@ Settings: relative to window 1
12:custom_tool_call_output:Script completed
Wall time <DURATION> seconds
Output:
| {"descriptions":["View a local image file from the filesystem when visual inspection is needed. Use this...>>; };\\n```"],"echo":"ECHOING: ping","agent":"/root"} [hash=6f5fd573af132df1]
| {"descriptions":["View a local image file from the filesystem when visual inspection is needed. Use this...>>; };\\n```"],"echo":"ECHOING: ping","agent":"/root"} [hash=dafcf2a1f1aff990]
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (13; hash=75ef17d8adc1b3db):
tools (13; hash=90319af6484fda67):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -27,7 +27,7 @@ Settings:
- function/curr_time
- function/sleep
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/collaboration: Tools for spawning and managing sub-agents.; hash=2c8d6295978dbe63
- function/followup_task
- function/interrupt_agent
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (14; hash=c23674e45d0ff686):
tools (14; hash=e60a1a331b573432):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -27,7 +27,7 @@ Settings:
- function/curr_time
- function/sleep
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/collaboration: Tools for spawning and managing sub-agents.; hash=2c8d6295978dbe63
- function/followup_task
- function/interrupt_agent
@@ -67,7 +67,7 @@ Settings:
## Window 2 (after request 2)
Settings: relative to window 1
tools (13; hash=75ef17d8adc1b3db):
tools (13; hash=90319af6484fda67):
- removed namespace/mcp__enterprise
-- request 3 (turn) --
00:message/developer:
@@ -15,8 +15,8 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (7; hash=dac6a853a6177003):
- custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=9ad48dbb5eac1f9b
tools (7; hash=83df59b286cda93b):
- custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=2051847a9a5d026d
- function/wait: Waits on a yielded `exec` cell and returns new output or completion. - Use `wait` only after `exe...; args=[cell_id, max_tokens, terminate, yield_time_ms]; hash=6692d8f6c55be489
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- function/request_user_input_async: Ask the user one or more questions during ongoing work. Use this tool only to request missing inf...; args=[questions]; hash=d6d74f33db330162
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (11; hash=e4ad6947d6c88cac):
tools (11; hash=a079637661e48f3a):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/multi_agent_v1: Tools for spawning and managing sub-agents.; hash=2327142bda177c50
- function/close_agent
- function/resume_agent
@@ -50,7 +50,7 @@ Settings:
## Window 2 (after request 2: input diverged at item 03)
Settings: relative to window 1
prompt_cache_key: "<PROMPT_CACHE_KEY 2>"
tools (11; hash=b0f271ad4e160a41):
tools (11; hash=ab95a443abed962b):
~ namespace/mcp__shared: Use these tools to exercise the rmcp test server.; hash=a9e155b340098fac
- function/echo_tool
-- request 3 (turn; joiner) --
@@ -15,8 +15,8 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (4; hash=c0d891b72dfe19c6):
- custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=5f9adb37c23e2085
tools (4; hash=d137abed31032f6c):
- custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=8f357b6218c02911
- function/wait: Waits on a yielded `exec` cell and returns new output or completion. - Use `wait` only after `exe...; args=[cell_id, max_tokens, terminate, yield_time_ms]; hash=6692d8f6c55be489
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- web_search; hash=e214d4db1bc49e89
@@ -32,8 +32,8 @@ Settings:
## Window 2 (after request 1)
Settings: relative to window 1
tools (4; hash=f74ded0dd31e19a8):
~ custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=f36b0269278fcd9d
tools (4; hash=f5a01fb2c9761a26):
~ custom/exec: Run JavaScript code to orchestrate/compose tool calls - Evaluates the provided JavaScript code in...; hash=dfbda3722b16bdf9
-- request 2 (request) --
00:message/developer:
<MODEL_INSTRUCTIONS>
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (11; hash=d2c3940ebd5ac02b):
tools (11; hash=2b93aeb57ddb3eb5):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/multi_agent_v1: Tools for spawning and managing sub-agents.; hash=2327142bda177c50
- function/close_agent
- function/resume_agent
@@ -45,7 +45,7 @@ Settings:
## Window 2 (after request 1)
Settings: relative to window 1
tools (11; hash=21579d54e22ad527):
tools (11; hash=8f2ae06e5a29e3e1):
~ namespace/mcp__discovery: Use these tools to exercise the rmcp test server.; hash=66db79539eaf6be3
- function/echo
- function/search_tools
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=2b84850bc279e5ed):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=83697a0118230669):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (2; hash=d56ea226c3637559):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (2; hash=5ea7a55416afee55):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -6,8 +6,8 @@ Scenario: A worker inherits the parent prefix unchanged and receives its own ins
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (2; hash=cefa72e421dc276f):
- namespace/functions; hash=125a3a92f05f72e9
00:additional_tools/developer (2; hash=855bf977570202a9):
- namespace/functions; hash=2e61665749f6eed1
- custom/exec
- function/wait
- function/request_user_input
@@ -46,8 +46,8 @@ Scenario: A worker inherits the parent prefix unchanged and receives its own ins
## Window 2 (after request 3: input diverged at item 12)
-- request 4 (turn) --
00:additional_tools/developer (2; hash=cefa72e421dc276f):
- namespace/functions; hash=125a3a92f05f72e9
00:additional_tools/developer (2; hash=855bf977570202a9):
- namespace/functions; hash=2e61665749f6eed1
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (2; hash=7b4c6a992fd91f91):
- namespace/functions; hash=101d32aa59ea258b
00:additional_tools/developer (2; hash=ba36cb129fbde5df):
- namespace/functions; hash=106dd1b983efd2f7
- function/exec_command
- function/write_stdin
- function/request_user_input
@@ -93,8 +93,8 @@ Settings:
## Window 2 (after request 3: input repeated)
Settings: same as window 1
-- request 4 (turn) --
00:additional_tools/developer (2; hash=7b4c6a992fd91f91):
- namespace/functions; hash=101d32aa59ea258b
00:additional_tools/developer (2; hash=ba36cb129fbde5df):
- namespace/functions; hash=106dd1b983efd2f7
- function/exec_command
- function/write_stdin
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=696793f2090ba3e3):
- namespace/functions; hash=eb439da1f745d8d5
00:additional_tools/developer (3; hash=ff4c5c6769e3f0a3):
- namespace/functions; hash=d81b6bdfabfd18f7
- custom/exec
- function/wait
- function/request_user_input
@@ -15,11 +15,11 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (6; hash=dde875513b2c7cff):
tools (6; hash=9a4caaeb38e57991):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/collaboration: Tools for spawning and managing sub-agents.; hash=3cbf715dc154e1b7
- function/followup_task
- function/interrupt_agent
@@ -6,8 +6,8 @@ Scenario: With retention disabled, the client can still send retain=true. Tool o
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -45,8 +45,8 @@ Scenario: With retention disabled, the client can still send retain=true. Tool o
## Window 2 (after request 3: input diverged at item 11)
-- request 4 (compaction) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -84,8 +84,8 @@ Scenario: With retention disabled, the client can still send retain=true. Tool o
## Window 3 (after request 4: input diverged at item 06)
-- request 5 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -6,8 +6,8 @@ Scenario: Retained thread instructions stop emergency tool-output trimming. Comp
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -48,8 +48,8 @@ Scenario: Retained thread instructions stop emergency tool-output trimming. Comp
## Window 2 (after request 4: input diverged at item 06)
-- request 5 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=5e8370540461db70):
- namespace/functions; hash=7d26e086c1dbba28
00:additional_tools/developer (3; hash=90063690a488dd0c):
- namespace/functions; hash=596780ea93468a82
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=5e8370540461db70):
- namespace/functions; hash=7d26e086c1dbba28
00:additional_tools/developer (3; hash=90063690a488dd0c):
- namespace/functions; hash=596780ea93468a82
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -160,8 +160,8 @@ Settings:
## Window 2 (after request 7: input diverged at item 02)
Settings: same as window 1
-- request 8 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -295,8 +295,8 @@ Settings: same as window 1
## Window 3 (after request 11: input diverged at item 07)
Settings: same as window 1
-- request 12 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -436,8 +436,8 @@ Settings: same as window 1
## Window 4 (after request 18: input diverged at item 10)
Settings: same as window 1
-- request 19 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -6,8 +6,8 @@ Scenario: Astra reads Cloud instructions, discovers an omitted deployment instal
## Window 1
-- request 1 (request) --
00:additional_tools/developer (3; hash=e075a43448be3246):
- namespace/functions; hash=bbcd688349ca7e12
00:additional_tools/developer (3; hash=c989fc4f95fb6f22):
- namespace/functions; hash=80f0e6670cf933cc
- custom/exec
- function/wait
- function/request_user_input
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (10; hash=df5c5997ff843305):
tools (10; hash=9f8074d045f3bd93):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- tool_search: # Tool discovery Searches over deferred tool metadata with BM25 and exposes matching tools for t...; args=[limit, query]; hash=b573ed69c0b0e6c7
- web_search; hash=f103e327c6870a46
-- request 1 (turn) --
@@ -7,8 +7,8 @@ Scenario: Available file references pass through. A later request replaces missi
## Window 1
-- request 1 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -43,8 +43,8 @@ Scenario: Available file references pass through. A later request replaces missi
## Window 2 (after request 1: input diverged at item 06)
-- request 2 (turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (11; hash=b18d58517f1e1d66):
tools (11; hash=f0e5643a27e61b18):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/mcp__cached_rmcp: Tools in the mcp__cached_rmcp namespace.; hash=fa6836ec918ed91a
- function/calendar_list_events
- namespace/mcp__codex_apps__calendar: Plan events and manage your calendar.; hash=78fbd295e219be87
@@ -141,7 +141,7 @@ Settings:
## Window 2 (after request 1)
Settings: relative to window 1
tools (11; hash=ab44f83a0745a3ed):
tools (11; hash=4235ff36fba52467):
~ web_search; hash=e214d4db1bc49e89
-- request 2 (turn) --
00:message/developer:
@@ -15,12 +15,12 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (7; hash=dba1108606ea8676):
tools (7; hash=0dea9012dd51867e):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/web: Tools in the web namespace.; hash=c60e4db83b9fd6b0
- function/run
- tool_search: # Tool discovery Searches over deferred tool metadata with BM25 and exposes matching tools for t...; args=[limit, query]; hash=b573ed69c0b0e6c7
@@ -15,7 +15,7 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (9; hash=5429df93e52f82a6):
tools (9; hash=1439533a92960b54):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/list_mcp_resources: Lists resources provided by MCP servers. Resources allow servers to share data that provides cont...; args=[cursor, server]; hash=6792edaff13561bd
@@ -23,7 +23,7 @@ Settings:
- function/read_mcp_resource: Read a specific resource from an MCP server given the server name and resource URI.; args=[server, uri]; hash=24246896b39008ae
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- tool_search: # Tool discovery Searches over deferred tool metadata with BM25 and exposes matching tools for t...; args=[limit, query]; hash=b573ed69c0b0e6c7
-- request 1 (turn) --
00:message/developer:
@@ -78,7 +78,7 @@ Settings:
## Window 2 (after request 7)
Settings: relative to window 1
model: "native-paper"
tools (10; hash=89fc167132311088):
tools (10; hash=29176ecf1284679a):
+ namespace/web: Tools in the web namespace.; hash=0037b1ed8d50fab5
- function/run
-- request 8 (turn) --
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn; Initial policy A) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -132,8 +132,8 @@ Settings:
## Window 2 (after request 4: input diverged at item 02)
Settings: unavailable (items only)
-- request 5 (items; Stored replacement before next turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -229,8 +229,8 @@ Settings: unavailable (items only)
## Window 3 (after request 5)
Settings: same as window 1
-- request 6 (turn; Unchanged policy C after compaction) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -16,8 +16,8 @@ Settings:
text: {"verbosity":"low"}
tool_choice: "auto"
-- request 1 (turn; Initial policy A) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -123,8 +123,8 @@ Settings:
## Window 2 (after request 4: input diverged at item 02)
Settings: unavailable (items only)
-- request 5 (items; Stored replacement before next turn) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -218,8 +218,8 @@ Settings: unavailable (items only)
## Window 3 (after request 5)
Settings: same as window 1
-- request 6 (turn; Unchanged policy C after compaction) --
00:additional_tools/developer (3; hash=bf2e30605ff5dc17):
- namespace/functions; hash=fada4a476a3754dd
00:additional_tools/developer (3; hash=e019c949fc483f83):
- namespace/functions; hash=03340277aa23936f
- custom/exec
- function/wait
- function/request_user_input
@@ -15,13 +15,13 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (9; hash=92200068832c7bdc):
tools (9; hash=8e4f89cf37119b4e):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/wait_for_environment: Wait for the selected environment.; args=[environment_id]; hash=aac8552da86b3b20
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/skills: Tools in the skills namespace.; hash=998a720eb0f017bc
- function/list
- function/read
@@ -15,13 +15,13 @@ Settings:
stream: true
text: {"verbosity":"low"}
tool_choice: "auto"
tools (9; hash=92200068832c7bdc):
tools (9; hash=8e4f89cf37119b4e):
- function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
- function/write_stdin: Writes characters to an existing unified exec session and returns recent output.; args=[chars, max_output_tokens, session_id, yield_time_ms]; hash=7fb62d8a3e2df4e2
- function/wait_for_environment: Wait for the selected environment.; args=[environment_id]; hash=aac8552da86b3b20
- function/request_user_input: Request user input for one to three short questions and wait for the response. This tool is only ...; args=[questions]; hash=64a729b25c424fc6
- custom/apply_patch: The `apply_patch` tool can be used to edit files. This is a FREEFORM tool, so do not wrap the pat...; hash=dea9f12b0111a4ba
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[detail, path]; hash=3b766d5bdec5d50e
- function/view_image: View a local image file from the filesystem when visual inspection is needed. Use this for images...; args=[path]; hash=1908de9c0be7118a
- namespace/skills: Tools in the skills namespace.; hash=998a720eb0f017bc
- function/list
- function/read
@@ -52,7 +52,7 @@ Settings:
## Window 2 (after request 1)
Settings: relative to window 1
tools (9; hash=92200068832c7bdc):
tools (9; hash=8e4f89cf37119b4e):
~ function/exec_command: Runs a command in a PTY, returning output or a session ID for ongoing interaction.; args=[cmd, justification, login, max_output_tokens, prefix_rule, sandbox_permissions, shell, tty, workdir, yield_time_ms]; hash=c04108e51a8548c5
-- request 2 (turn) --
00:message/developer:
+1 -1
View File
@@ -585,7 +585,7 @@ async fn mcp_image_output_preserves_image_and_no_text_summary() -> Result<()> {
);
assert_eq!(
arr[1],
json!({"type": "input_image", "image_url": openai_png, "detail": "high"})
json!({"type": "input_image", "image_url": openai_png, "detail": "original"})
);
Ok(())

Some files were not shown because too many files have changed in this diff Show More