openai/codex
1
0
Fork 0
mirror of https://github.com/openai/codex.git synced 2026-09-13 11:47:17 +00:00
Code Issues Packages Projects Releases Wiki Activity

agentydragon(tasks): fix numbering in Implementation steps

Browse Source
This commit is contained in:
Rai (Michael Pokorny)
2025-06-24 19:01:54 -07:00
parent 8a6a6dc3fb
commit b618d3a8ea
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
agentydragon/tasks/02-auto-approve-predicates.md
View File

@@ -32,8 +32,8 @@ Multiple scripts cast votes: if any script returns `deny`, the command is denied
- On `allow`, auto-approve under sandbox.
- On script errors or unrecognized outputs, immediately prompt the user (via `request_command_approval`) with a reason string describing the error/output, then run or reject based on their decision.
- On `no-opinion`, fall through to the next script.
6. Write async unit tests for `get_auto_allow_vote`, covering allow, deny, no-opinion, and error-exit cases.
7. Update documentation (`config.md`) to document the new auto-approval predicates feature.
5. Write async unit tests for `get_auto_allow_vote`, covering allow, deny, no-opinion, and error-exit cases.
6. Update documentation (`config.md`) to document the new auto-approval predicates feature.
## Notes
- This pairs with the existing `approval_policy = "unless-allow-listed"` but now ensures script errors or unexpected outputs trigger a targeted manual approval prompt with context.