refactor: standardize Codex agent execution pattern#70
Merged
arittr merged 4 commits into750f25-task-1-extract-prompt-modulefrom Nov 3, 2025
Merged
Conversation
Refactor Codex agent to use direct CLI execution with stdin (matching Claude/Gemini pattern) instead of temp file I/O, reducing complexity and improving consistency.
Changes:
- Updated executeCommand() to use exec('codex', ['exec', '--skip-git-repo-check'], { input: prompt }) pattern
- Removed temp file logic: _readOutput(), _cleanupTempFile() methods (~50 LOC reduction)
- Codex agent now ~110 LOC (down from ~160 LOC)
- Updated tests to reflect new execution pattern (no temp file mocking)
- All three agents (Claude, Codex, Gemini) now have identical test structure
Acceptance Criteria Met:
- ✅ Codex uses stdin pattern (no temp files)
- ✅ File handling methods removed (~50 LOC reduction)
- ✅ Codex agent simplified to ~110 LOC
- ✅ Tests updated to stdin pattern
- ✅ All agents have identical test structure
Owner
Author
|
This change is part of the following stack: Change managed by git-spice. |
This was referenced Nov 3, 2025
Remove dual-mode complexity by eliminating manual fallback mode. Add quiet flag for suppressing progress messages. Changes: - Removed --no-ai flag from CLI and schemas - Added --quiet flag for suppressing stderr output - Removed enableAI config option from generator - Removed manual fallback methods from generator (~150 LOC) - Generator always uses AI; agent failures throw errors - Default behavior shows generating message in all modes - Error messages provide clear installation instructions Acceptance Criteria Met: - ✅ --no-ai flag removed - ✅ --quiet flag added and working - ✅ enableAI config removed - ✅ Manual fallback methods removed (~150 LOC) - ✅ Generator always uses AI - ✅ Default shows progress message - ✅ Clear error messages - ✅ Net ~200 LOC reduction 🤖 Generated with Claude via commitment Co-Authored-By: Claude <noreply@anthropic.com>
Document breaking changes with clear migration instructions. Update all docs to reflect new AI-only architecture. - Added CHANGELOG section for breaking changes and migration guide - Provided installation links for Claude, Codex, Gemini CLIs - Removed all --no-ai references from README - Documented --quiet flag with examples - Updated CLAUDE.md to reflect AI-only architecture and prompts module - All documentation accurate and up-to-date Acceptance Criteria Met: - ✅ CHANGELOG documents breaking change with migration - ✅ CHANGELOG includes CLI installation links - ✅ README removes --no-ai references - ✅ README documents --quiet flag - ✅ CLAUDE.md reflects new architecture - ✅ All documentation accurate
Fix mock pollution in shell integration tests by using cache-busting dynamic imports. Issue: Agent unit tests use mock.module() which creates persistent module mocks. Even after mock.restore(), Bun's module cache still serves the mocked version to subsequent imports. Solution: - Dynamically import shell module in beforeAll hook after calling mock.restore() - Add timestamp query parameter to force fresh import and bypass module cache - Rename test file to zzz-shell.integration.test.ts for consistent ordering Result: All 14 failing tests now pass - Before: 568 pass, 14 fail - After: 582 pass, 0 fail Tests verified: bun test completes successfully
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Refactor Codex agent to use direct CLI execution with stdin (matching Claude/Gemini pattern) instead of temp file I/O, reducing complexity and improving consistency.
Changes:
Acceptance Criteria Met: