clone/codex
1
0
Fork 0
mirror of https://github.com/openai/codex.git synced 2026-04-24 14:45:27 +00:00
Code Issues Packages Projects Releases Wiki Activity

agentydragon(prompts): add Commit agent & update Developer agent to run pre-commit before handoff

Browse Source
This commit is contained in:
Rai (Michael Pokorny)
2025-06-25 00:47:25 -07:00
parent 6059f47a79
commit 2c1c65a004
3 changed files with 17 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
agentydragon/README.md
View File

@@ -59,8 +59,9 @@ Tasks live under `agentydragon/tasks/` as individual Markdown files. Please upda
-
- Without `--agent`, this creates or reuses a worktree at
- `agentydragon/tasks/.worktrees/<task-id>-<task-slug>` off the `agentydragon` branch.
- With `--agent`, after setup it launches a Codex agent in that workspace,
- With `--agent`, after setup it launches a Codex agent in that workspace,
- preloaded to work on the specified task from its Markdown file.
- After the Developer agent finishes and updates the task file, the Commit agent will automatically stage and commit the changes with an appropriate commit message. You do not need to run `git commit` manually.
---

12
agentydragon/prompts/commit.md Normal file
View File

@@ -0,0 +1,12 @@
## Commit Agent Prompt
You are the **Commit** Codex agent for the `codex` repository. Your job is to stage and commit the changes made by the Developer agent.
Do **not** modify any files; only perform Git operations to record the work done.
When you run:
- Stage all modified files.
- Run the repository’s pre-commit hooks on these changes (e.g. `pre-commit run --files <changed-files>`); if any hooks modify files, stage those fixes as well.
- Commit with a message prefixed `agentydragon(tasks):` followed by a concise summary of the work performed as described in the task’s **Implementation** section.
- Stop and await further instructions.
Do not edit any code or Markdown files. Only run Git commands to finalize the Developer agent’s work.

5
agentydragon/prompts/developer.md
View File

@@ -15,6 +15,7 @@ only ask clarifying questions if you encounter genuine ambiguities in the requir
At any point, you may set the task’s **Status** to any valid state (e.g. Not started, Started, Needs input, Needs manual review, Done, Cancelled) as appropriate. Use **Needs input** to request further clarification or resources before proceeding.
When you have completed the implementation and updated the task file:
- set the task’s **Status** to "Done".
- stage and commit all changes (including your code and the updated Markdown) with a commit message prefixed `agentydragon(tasks):`, summarizing the work performed. You **MUST** commit these changes **BEFORE** pausing or stopping.
set the task’s **Status** to "Done".
- Run the repository’s pre-commit hooks on all changed files (e.g. `pre-commit run --files <changed-files>`), and stage any autofix changes.
- Do **not** stage or commit beyond hook-driven fixes. Instead, stop and await the Commit agent to record your updates.
Then stop and await further instructions.