## What changed
- Register both `clock.curr_time` and `clock.sleep` when a model lists `clock` in `experimental_supported_tools`.
- Keep current-time reminder injection controlled separately by the `current_time_reminder` feature.
## Testing
- Cover a model-advertised clock with reminders disabled, verifying that both tools are available and no reminder is added.
GitOrigin-RevId: c39b5cb6e2da397daaaa4e0ad81f0763b8dd91e9
## Why
Read-deny policies must use the target executor's path convention so URI-based
policy checks and native filesystem enumeration enforce the same rules.
## What changed
- Prepare deny roots and glob matchers from `PathUri` policy context, including
executor-relative working directories and home-relative patterns.
- Match Windows globs case-insensitively with normalized separators, while
preserving byte-oriented POSIX matching for non-UTF-8 paths.
- Fail closed for malformed paths, incompatible path conventions, unresolved
home-relative patterns, and invalid globs.
- Make Windows deny-read discovery use case-insensitive ripgrep glob matching.
## Testing
Added coverage for Windows URI conventions, executor home expansion,
case-insensitive `.env` discovery, malformed paths, non-UTF-8 names, and
canonical directory-link targets.
GitOrigin-RevId: 36001219a2e9b36acfce8972dc1d3bbc304271c5
## What changed
- Load the effective plugin configuration for every requested `cwd` in
`plugin/list` and `plugin/installed`.
- Combine local marketplaces in request order, keeping the first source for
duplicates while merging installed and enabled state across repositories.
- Report invalid project configurations through `marketplaceLoadErrors` without
dropping valid local catalogs or globally enabled remote catalogs.
- Use the combined marketplace context for forced and background cache refreshes,
while keeping remote plugin feature gating request-wide.
## Testing
- Cover distinct and duplicate marketplaces, cross-repository enabled state,
invalid project configuration, local cache refresh, and global versus
project-local feature flags.
GitOrigin-RevId: 53c8e0e9c760dc2b969ab10d0e63fb439673c230
## What changed
- Add the optional `platformOs` field to exec-server environment metadata, with
local exec servers reporting `std::env::consts::OS` and legacy servers
remaining compatible when the field is absent.
- Preserve the reported OS in each selected turn environment, including the
local environment.
## Testing
- Extend protocol serialization and environment-selection coverage for remote
and local platform values.
GitOrigin-RevId: 0f325f6cb83cd8f8c7ec1ce2a4041f45a29e8f38
## What changed
- Add an optional `multi_agent_reasoning_effort` field to model metadata.
- When `Ultra` is selected, use a supported catalog override when present. Otherwise, prefer `max`, then the highest supported non-`ultra` effort, and finally `medium` when the model has no reasoning levels.
- Preserve existing request behavior for other reasoning efforts, including mapping `persistent` to `disabled`.
## Testing
- Cover configured, missing, unsupported, and empty-catalog fallbacks for both parent and spawned-agent requests.
GitOrigin-RevId: b7ef576f32e0ea95ef38bf95c25789ed16aae94e
## What changed
- Count remote executor MCP discovery attempts, including refreshes, with `success` and `error` outcomes.
- Count each discovered server by name and report whether it was found or unavailable before policy and startup filtering.
GitOrigin-RevId: b912aa64b606e6243cce79df0fe85d485a2c3547
## What changed
- Report the executor's user home in environment metadata.
- Cache the reported home on selected turn environments and include it in
filesystem sandbox contexts, including `apply_patch` requests.
- Preserve compatibility with executors that do not report a home directory.
## Testing
- Cover local, remote, inherited, serialized, and `apply_patch` sandbox
contexts.
GitOrigin-RevId: ea65e68e33a80bc9bcfc7c53b1486ab944699470
## What changed
- Add `ToolLifecycleContributor::on_mcp_tool_result` with access to the
executed MCP tool context, rewritten arguments, extension data stores, and
mutable server result.
- Run contributors before publishing MCP completion and before preparing the
result for the model, so extensions can inspect or replace successful and
error results.
## Testing
- Cover unchanged, replaced, and error results for direct MCP calls, plus
replaced and error results through Code Mode.
- Verify completion waits for result processing and the processed result flows
to both completion events and subsequent model input.
GitOrigin-RevId: ece7acfe9c5eed4778cb78a29c3ffd06a74e1bcf
## What changed
- Add `mcp_optional_startup_grace_ms` with a default of 1,000 ms to control how long tool catalog capture waits for optional MCP servers.
- Treat a value of `0` as disabling the shared grace, so optional servers use their configured `startup_timeout_sec` instead.
- Apply updated grace values during runtime and MCP configuration refreshes, and reset cached startup deadlines when the configured duration changes.
## Testing
- Cover custom, disabled, refreshed, and shared startup grace behavior in MCP connection and core integration tests.
GitOrigin-RevId: 936196215ed6e0a29c9ec72cfee17663978df8ef
Expand Windows deny-read globs robustly across protected directories while preserving filesystem enforcement. Preserve structured MCP tool and resource errors through app-server responses. Bind cached Guardian classifications to current local and root authorization state so stale scores cannot approve actions.
Add regression coverage for the sandbox resolver, structured protocol errors, and authorization changes.
GitOrigin-RevId: 4b80ed724d869afeca79204222d8465fa99d3a24
## Why
Planning tools for a candidate model must not change the selected model or its
Responses Lite tool inventory. Otherwise, preparing a fallback can overwrite
metadata before the current model's request is sent.
## What changed
- Build tool registries and visible specs from an explicit `ModelInfo`.
- Store the finalized tool mode, code-mode names, namespace inventory, and
terminal and child-management capabilities on `ToolRouter`.
- Publish namespace inventory only after selecting the step's router, and use
that router's mode when dispatching and recording tool calls.
## Testing
Added coverage for candidate-model isolation, fallback compaction metadata,
code-mode worker selection, normalized tool names, and router capabilities.
GitOrigin-RevId: 56df2de8f42c7e93a4c4dcd117720e40a0cca3db
## Why
Test fixture setup can take more than five seconds on loaded CI workers, and
failures returned from setup can be obscured when downstream mocks verify their
expectations during teardown.
## What changed
- Add a shared `expect_startup` helper with a 30-second timeout that panics on
either timeout or setup failure before mock teardown runs.
- Use the helper for remote-environment and unified-exec fixture setup, and
allow the same startup window for the initial exec-server requests.
## Testing
Add paused-time tests covering successful startup after five seconds, timeout
reporting, and propagation of setup errors in the presence of unmet mocks.
GitOrigin-RevId: 39db69bde2983acd023a1911a6fad106674e169a
## Why
Remote plugin sync results need to identify which runtime categories may require
invalidation when a cached bundle changes.
## What changed
- Add `RemotePluginCapabilities` hints for MCP servers, Apps, hooks, and skills.
- Attach capability hints to materialized and removed plugins. Updates union the
declarations from the old and new bundles, while removals inspect the active
bundle before deleting its cache.
- Discover declarations without applying runtime policy or creating runtime
data directories.
## Testing
- Cover Agent Plugin capability discovery with an unavailable runtime data
directory.
- Cover capability accumulation across bundle versions and removal reporting.
GitOrigin-RevId: 338075d22d86f30f52be5785a2d87c77b33609c6
## Why
Restored app-server threads can carry permission profiles that cannot be
represented without loss as legacy sandbox-mode overrides. Sending those
profiles as per-turn overrides can discard their original permission details.
## What changed
- Distinguish restored permission profiles from legacy per-turn sandbox
overrides.
- Preserve restored profiles while routing turns and changing to compatible
working directories.
- Reject `/cd` when a restored profile cannot be represented safely by a
legacy sandbox mode.
## Testing
- Cover retaining a restored read-only profile across `/cd` without a turn
override.
- Verify that lossy sandbox-mode projections are detected.
GitOrigin-RevId: 6143dca4b83ab4f59bab05f6a59b4fe841c56192
## Why
Guardian tests could continue after a WebSocket handshake but before the
connection entered the idle pool. The shared request waiter could also miss a
notification delivered between checking the request log and registering its
wait.
## What changed
- Wait for Guardian prewarming to populate the idle connection pool instead of
treating completed handshakes as sufficient.
- Use two initial Guardian connections in tests while retaining eight in
non-test builds.
- Register for request-log notifications before inspecting the log to avoid a
lost wakeup.
GitOrigin-RevId: 88f6a22c5401e03c24e6e114f2563fc641bfef80
## What changed
Emit a `codex.unified_exec.stdin_review.size_check` counter when reviewing
input for an escalated terminal. Tag each check with:
- the result: `formatter_truncated`, `over_limit`, or `within_limit`;
- whether the submitted input is control-only or text.
Preserve the existing rejection behavior for truncated or oversized reviews.
GitOrigin-RevId: 6556beb2121cfaf91d572a26822b50442391c95f
## What changed
- Roll token usage from spawned descendants, including nested subagents, into the root goal's usage.
- Apply descendant usage during active and idle progress accounting so it contributes to token budgets.
- Reset descendant accounting baselines when the active goal changes and preserve usage recorded concurrently with a checkpoint.
## Testing
- Cover child and grandchild usage, budget exhaustion, unloaded parent runtimes, goal replacement, idle accounting, and concurrent checkpoints.
GitOrigin-RevId: 8f97ec6778c55b9adf94b887b5fd03999ed0eb94
## What changed
- Tell the `spawn_agent` tool to set `model` only when the user explicitly
requests a different model.
- Update the tool-description test to enforce the stricter guidance.
GitOrigin-RevId: ed46bcb544876e408896e0d8c54292192369951c
## Why
Model settings can change between steps in the same turn. Token-budget defaults,
context-window limits, and model-visible guidance need to stay aligned with the
model captured for each step.
## What changed
- Preserve the turn's original token-budget preferences and resolve model-owned
defaults when each `StepContext` is created.
- Use the step's model to determine context-window availability and token-budget
guidance, while keeping explicit user settings unchanged.
- Emit one-time replacement or removal notices when context-window guidance
changes, including compatibility with previously persisted world-state
snapshots.
- Add `ModelInfo::usable_context_window()` to distinguish reserved-headroom
capacity from the resolved context window and auto-compaction limit.
## Testing
- Cover mid-turn model switches with model defaults, explicit settings, config
reloads, missing context windows, and missing destination guidance.
- Cover unchanged, replaced, removed, blank, and legacy guidance transitions.
GitOrigin-RevId: 91100a9f3625bef8a79ecd158eb028051c309eee
## Why
Reviewing a shortened `write_stdin` action while sending the complete input could allow unreviewed bytes to reach an escalated terminal.
## What changed
- Require the fully formatted action and approval reason to fit within 8,000 bytes.
- Reject truncated or oversized actions before requesting approval or writing any bytes to the terminal.
- Document the reviewed stdin size limit in the app-server protocol.
## Testing
Add an integration test that uses JSON escaping to push a formatted action over the limit and verifies that it is neither offered for approval nor executed, while a subsequent valid write still succeeds.
GitOrigin-RevId: f6c96a78a65a136366c8119a7d242dfc2ec05fe6
## What changed
- Lower the default `max_tool_call_lag` from three tool calls to two, causing
older async risk scores to fall back to strict review sooner.
- Update the async scorer integration test to verify score expiration using the
default lag instead of a test-specific override.
GitOrigin-RevId: a07cae3634428a1e64745674eced92fb770f441a
## Why
Guardian classification must not proceed after silently dropping a parent
compaction that cannot fit within its configured size limit.
## What changed
- Distinguish an absent or unusable parent compaction from serialization and
size-limit failures.
- Record a fail-closed score and failed classification when the latest encrypted
parent compaction cannot be serialized or exceeds the complete-item byte
budget.
- Preserve omission of compactions with missing or empty encrypted content
without falling back to older context.
## Testing
Cover compactions at the exact byte limit, one byte over it, and oversized
passthrough metadata.
GitOrigin-RevId: 8e607c88f10003cdfab273963bc73c848b09c821
## What changed
- Move bounded planned-action JSON rendering from the Guardian V2 extension implementation into a dedicated `async_scorer::action` module.
- Keep the structural size-limit tests alongside the extracted implementation.
GitOrigin-RevId: 9292b164588ba0119b4c5ff81056bf7416352d69
## What changed
- Store per-turn trusted skill paths directly in a `BTreeSet` instead of a
mutex-protected set.
- Mutably record paths and consume the collection when producing the final
ordered path list.
- Strengthen the bounded-deduplication test to verify the exact collected paths.
GitOrigin-RevId: 69e87fc7783738d85ed204f18a3968d0837c3693
## What changed
Represent completed and superseded Guardian V2 classifications with a dedicated
`ClassificationOutcome` enum. Map the typed outcome to the existing metrics
status and emit truncation metrics only for scored classifications.
GitOrigin-RevId: b00d345cb0162ed3fc7f53632630b631f18820e7
## Why
Large or minified diffs can require substantial work to allocate wrapped rows and syntax-highlighted spans in the inline patch view.
## What changed
- Limit inline patch previews to 12 rendered rows across all changed files, while retaining file summaries and directing users to the transcript when content is omitted.
- Preserve complete diff content in transcript and raw views.
- Bound preview source scanning to 64 KiB so zero-width content cannot bypass the row budget.
## Testing
Add coverage for added, deleted, updated, renamed, minified, narrow-width, wide-character, combining-character, and multi-file diffs, including preservation of complete transcript and raw output.
GitOrigin-RevId: 1899129ab548771c29d226ff624549573fb2f688
## Why
Guardian reviews for delegated workers need to recognize skills explicitly
invoked by the user on the root turn.
## What changed
- Record verified user-owned skill paths as bounded, turn-matched Guardian
evidence.
- Include trusted paths from the current worker and its root snapshot when
building review context.
- Exclude skill trust from later root turns so stale authorization is not
inherited.
## Testing
- Add an app-server lifecycle test covering inheritance of a root-invoked
skill by a delegated worker and its removal on the next root turn.
GitOrigin-RevId: aa28f485edb7a447e51d51de8dc89b959fd63d40
## Why
Plugin MCP trust decisions must describe the exact host-discovered server
registration. Looking up a plugin again by ID can resolve a different root.
## What changed
- Retain the host plugin root in MCP catalog attribution and connection identity.
- Reconnect an MCP server when its attributed plugin root changes.
- Pass the retained root through tool lifecycle attribution and use it directly
when building trusted tool context.
## Testing
Add coverage for catalog identity, connection reconciliation, lifecycle
attribution, and rejecting a different root associated with the same plugin ID.
GitOrigin-RevId: bc92743dd40ab44065a07c3ea972b4956ef8351f
## What changed
Define a shared 30-second timeout for Guardian V2 async tests and use it
for classifier requests, score updates, and authentication refresh waits.
GitOrigin-RevId: fa0e63e5e4c2b9ba0e3a09f0c8758886a867e546
## What changed
- Record classified risk levels in `codex.guardian_v2.classification.risk`.
- Record fast approval outcomes and their reasons in
`codex.guardian_v2.fast_decision`, including low-risk approvals and deferred
decisions caused by missing, stale, failed, out-of-scope, or elevated-risk
scores.
- Cover the new metric labels across approval and failure paths.
GitOrigin-RevId: e180054e0b3796e479a706157dd9ea7c9f03c001
## What changed
- Recognize `codex_sensitive_action: true` on MCP approval elicitations.
- Route marked actions to the synchronous Guardian reviewer instead of extension fast approval. Absent or false values preserve the existing path.
- Cover ordinary and strict auto-review configurations, including marked, unmarked, and explicitly false metadata.
GitOrigin-RevId: dce13d1696568ca81c8d2b08a7ef0d5f13e155bf
## What changed
- Parse optional `usage_metadata.amount` values from Responses API completion events without converting their string representation.
- Propagate the metadata through SSE and WebSocket streams, regular turns, and remote compaction.
- Add `usageMetadata` to the app-server `rawResponse/completed` notification and its generated schemas.
## Testing
- Cover missing, null, zero, and high-precision amounts, including completion events emitted during compaction.
GitOrigin-RevId: b16d81e8350eab0dc7ff613d0dc9dbed73041713
## What changed
- Add optional Browser Use and Computer Use confirmation-policy Markdown to model catalog messages.
- Send the configured documents verbatim in `openai/confirmation_policies` metadata for `node_repl` and `cua_repl` tool calls. Send an empty object when the issuing model has no policies so runtime defaults are cleared.
- Use the issuing step's model snapshot across approval waits, and omit the metadata for other MCP servers and Guardian sessions.
## Testing
- Cover policy serialization, configuration overrides, model changes, approval waits, partial or empty policies, eligible server filtering, and Guardian exclusion.
GitOrigin-RevId: 038c4c262a31fc71986e4a2a947fe9bcbd407acc
## What changed
Expand the `send_user_message_async` tool description to:
- identify questions, blockers, significant findings, and in-progress replies as
messages that may need immediate user attention;
- distinguish those messages from routine progress updates that belong in
commentary;
- clarify that sending a message neither ends the turn nor waits for a reply;
and
- encourage formatting questions so they are easy to notice and answer.
GitOrigin-RevId: 004ba7fd77b2e225a4baa2ab69d08660a2cf73d7
## What changed
- Serialize each history and notes request's output truncation policy into the
`x-openai-tool-output-truncation-policy` header.
- Forward the invoking tool call's policy for tool requests and use the thread
hint byte limit for context-contributor requests.
GitOrigin-RevId: 9d4e61480397b325efdb033861893e88c708de3b
## Why
Executed-tool metadata for a Code Mode cell can span its `exec` and `wait`
outputs. Consumers need to know whether that metadata contains the cell's full
tool call inventory or only a partial record.
## What changed
- Associate recorded tool calls with the originating Code Mode cell across
`exec` and `wait` outputs.
- Add a host-owned `tool_calls_complete` marker when recording finishes without
dropped or truncated calls. The marker describes inventory completeness, not
tool success.
- Preserve the cell metadata through replay and pruning, and clear the
completeness marker when prompt budgeting makes the record incomplete.
## Testing
Add coverage for multi-wait cells, retries, interruptions, recording limits,
argument truncation, prompt budgeting, and rejection of input-forged metadata.
GitOrigin-RevId: 377014edec06aab42aadab78b8b5dccc0a1c83ba
## What changed
- Add bundled proactivity and follow-up guidance when `ReasoningEffort::Persistent` is selected.
- Allow model metadata to override or disable the guidance with `persistent_instructions`.
- Track the instructions as world state so effort and model changes replace or retire stale guidance without duplicating unchanged context.
- Tailor approval guidance to `send_user_message_async` availability, rebuild it for forked agents and rolled-back history, and exclude Guardian sessions.
## Testing
- Cover default, overridden, disabled, replaced, and retired instructions.
- Cover model changes, history transitions, agent forks, Guardian sessions, and asynchronous-message tool availability.
GitOrigin-RevId: 78e05d0335a70b385bd069b45edcd4ba570cef91
## Why
Delegated prompts should retain the authority of the TUI tool that sent them instead of being recorded as user input.
## What changed
- Start delegated `create_thread` and `send_message_to_thread` turns with a namespaced tool output and an empty user input list.
- Recognize delegated outputs from trusted Codex namespaces when building activity summaries.
- Render replayed delegation outputs in history and transcripts with their source task attribution, without adding them to composer history.
## Testing
- Cover turn request encoding, namespace filtering, activity summaries, and replayed history/transcript rendering.
GitOrigin-RevId: aeb2f5c8ad2f2a898c80e8fc791ed879a0ab655d
## What changed
- Mark history and notes search queries, appended note text, and replacement note text as encrypted in their tool schemas.
- Send `x-openai-encrypted-tool-arguments: true` for the corresponding backend routes without changing the JSON request body.
## Testing
- Cover encrypted schema fields and backend headers for history search, notes search, append, and write calls.
- Verify unrelated history and notes requests do not receive the encrypted-arguments header.
GitOrigin-RevId: 58263cff832c979b436ddb4013b5c27218aca678
## Why
Trusted Guardian reviewer sessions can emit events without an app-server listener, leaving their turns and tool usage absent from analytics.
## What changed
- Translate Guardian session lifecycle and canonical tool item events into analytics notifications while excluding messages, reasoning, and review content.
- Preserve Guardian lineage during prewarming even when inherited client metadata is not yet available, without emitting an incomplete thread initialization event.
- Limit direct session tracking to trusted Guardian children so resumed threads that have an app-server listener do not report duplicate events.
## Testing
- Cover completed, failed, and interrupted Guardian turns, tool counts and timing, lineage, content exclusion, prewarmed lineage, and resumed-thread deduplication.
GitOrigin-RevId: 70ace5f7ed30a7e3b6a9d6097c8b2a6a326dc4f7
## What changed
- Add callback lifetimes to extension `ToolCall`, `ToolEnvironment`, turn-input context, and skill-read request types.
- Require extension tool executors to handle calls for any invocation lifetime and tie their returned futures to that lifetime.
- Update built-in extensions, adapters, and tests to use the lifetime-scoped APIs without retaining host-owned environment capabilities.
GitOrigin-RevId: aca9ce3a1b6870df550bc84ec9308cc65533f161
## What changed
- Inject W3C `traceparent` metadata into code-mode session and execution requests.
- Carry each execution's trace context on streamed tool callbacks so callback and nested-tool spans remain connected across the gRPC boundary.
- Preserve the relevant spans when execution, callback, and dispatch work moves into spawned tasks.
GitOrigin-RevId: 6482963cb9b2e2f0ea27f1345bba208b70a80d0b
## Why
Repeated skill locator roots can make catalog prompts larger even when the
catalog fits within its metadata budget.
## What changed
Evaluate aliased catalogs regardless of budget pressure and select them when
they preserve skill inclusion and description content while reducing prompt
size. Apply the same selection to combined host, executor, and orchestrator
catalogs.
## Testing
Add coverage for alias selection in full host-only catalogs and update skill,
symlink, and orchestrator resource tests to expect shortened locators and their
root mappings.
GitOrigin-RevId: ccedd3f3f6c7efaa1e5177a2d8c3d81d59239f52
## Why
Guardian treated all skill instructions as untrusted, so it could not use the
invocation of a user-owned skill as authorization evidence.
## What changed
- Record explicit and implicit skill invocations through extension contributors.
- Send Guardian a bounded, deduplicated list of canonical skill paths only when
they resolve under the user's Codex or `.agents` skill directories.
- Keep repository skills and symlinks that escape trusted roots untrusted, and
send only verified paths rather than skill contents to the classifier.
- Avoid recording host skill prompts that were superseded by core skill
injection.
## Testing
- Added unit coverage for trusted roots, symlink escapes, deduplication, and
size limits.
- Added integration coverage showing that Guardian trusts invoked user skills
while rejecting a forged repository skill.
GitOrigin-RevId: 1cc0be230f7e6649110201c23c3de8f26447ff36
## What changed
- Fetch ChatGPT account access and attach `cyber_trusted_access` under
`openai/entitlementContext` when an installed or selected plugin explicitly
requests it from a local, read-only stdio tool with no arguments.
- Reject caller-supplied entitlement context on direct MCP calls, and fall back
to an `unknown` status when access cannot be verified or the account changes
during the lookup.
- Count entitlement lookup time against the MCP tool-call timeout.
## Testing
- Cover eligible calls plus unowned, remote, HTTP, mutable, argument-bearing,
API-key, undeclared, direct, and account-switch cases in core and app-server
integration tests.
GitOrigin-RevId: 303b61585ef4530276c373d914a2d0ccef1f4ac2
## What changed
- Add an optional `toolOutput` to `turn/start` for starting or steering a turn with a named function-call output instead of user input.
- Emit standalone outputs as `functionCallOutput` thread items, persist them in durable history, and include them in resumed threads.
- Validate that tool outputs have a name, are not combined with nonempty `input`, satisfy the text-size limit, and use supported image URLs.
## Testing
- Cover standalone tool-output notifications, model input, and resumed history in both legacy and paginated history modes.
- Cover the text-size limit for structured tool output.
GitOrigin-RevId: a1eeec7b9b7de67b1a1c5e829a54552b85ec118b
## Why
Filesystem policies can describe paths using a convention that differs from the
host running Codex. Native path comparisons can therefore mis-handle cases such
as case-variant Windows paths or ambiguous encoded components.
## What changed
- Resolve policy entries and special roots as `PathUri` values using the
executor's path convention.
- Use validated URI components for containment, overlap, and precedence, and
fail closed when component boundaries are ambiguous.
- Restrict relative joins to descendants and apply the same matching rules to
protected metadata paths and permission-profile intersections.
## Testing
Added coverage for Windows case variants, encoded and opaque paths, repeated
separators, descendant joins, special roots, metadata protection, and preserved
deny entries.
GitOrigin-RevId: fb09d44d11df25faaac806fe00457e6f6b0d8596
## Why
Linux managed networking routed proxy connections through filesystem-backed Unix sockets, requiring an extra readable sandbox path and stale socket cleanup.
## What changed
- Transfer loopback TCP listeners from the sandbox namespace to host bridge processes over anonymous Unix socket pairs.
- Validate transferred descriptors, keep bootstrap channels private to each sandbox launch, and close them before running untrusted commands.
- Tie proxy bridge lifetime to the sandbox so cancellation closes active upstream connections.
## Testing
- Cover descriptor validation and interrupted handoffs.
- Verify concurrent sandboxes keep proxy endpoints isolated and do not inherit privileged sockets.
- Exercise managed networking end to end through `app-server`, confirming proxied access succeeds while direct host access remains blocked.
GitOrigin-RevId: 6561e09c83b927c604951bf2aa00ff532d179da5
## What changed
- Promote `compaction_image_budget` to stable and enable it by default.
- Charge retained images against the remote compaction context budget unless the feature is explicitly disabled, trimming older images as needed.
## Testing
- Update remote compaction coverage for the default, explicitly enabled, and explicitly disabled behavior.
GitOrigin-RevId: 0f715347cefc33b814bdcfd9b430ef453d3bd6df
