feat: update /init command to focus only on non-obvious discoveries #7400
Conversation
- Emphasize that only non-obvious information discovered by reading files should be included - Update all mode-specific examples to show truly non-obvious rules - Add CRITICAL section explaining what qualifies as non-obvious - Exclude standard patterns and framework defaults - Focus on gotchas, hidden requirements, and counterintuitive patterns
dosubot
bot
added
the
size:L
There was a problem hiding this comment.
Pull Request Overview
Updates the built-in /init slash command to generate more focused and valuable AGENTS.md documentation by emphasizing only non-obvious, project-specific information discovered through file analysis.
- Adds explicit criteria for what constitutes "non-obvious" information worth documenting
- Updates output structure to exclude obvious framework defaults and standard practices
- Transforms all mode-specific examples to focus on hidden gotchas and counterintuitive patterns
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
![roomote[bot]](https://avatars.githubusercontent.com/in/1546624?s=60&v=4){kind=small}
There was a problem hiding this comment.
Thank you for this valuable improvement! The focus on non-obvious, project-specific information will significantly enhance the quality of generated AGENTS.md files. I've reviewed the changes and have some suggestions to further improve clarity and consistency.
| 5. Document critical patterns | ||
| - Project-specific utilities | ||
| - Non-standard approaches | ||
| - Project-specific utilities (that you discovered by reading code) |
There was a problem hiding this comment.
Great addition of the CRITICAL section! The clarity on what constitutes non-obvious information is excellent. However, I noticed the terminology switches between "Discovered by reading files" (line 28) and "discovered by reading code" (line 59). Should we standardize on one phrase for consistency?
| - Database queries must use the query builder in packages/evals/src/db/queries/ | ||
| - Always use safeWriteJson() from src/utils/ instead of JSON.stringify for file writes | ||
| - Test coverage required for all new features in src/ and webview-ui/ | ||
| # Project Coding Rules (Non-Obvious Only) |
There was a problem hiding this comment.
These examples are very helpful! Quick question: Are these actual requirements in the Roo-Code codebase (like safeWriteJson() being mandatory), or are they hypothetical examples? If hypothetical, should we add a note clarifying they're illustrative?
| - Focus on gotchas, hidden requirements, and counterintuitive patterns | ||
| - Include specific file paths when referencing custom utilities | ||
| - Be extremely concise - if it's obvious, don't include it | ||
| - Every line should prevent a potential mistake or confusion |
There was a problem hiding this comment.
The quality criteria is excellent, especially the test question. Consider adding guidance for edge cases - what should AI assistants do when unsure if something is "non-obvious"? Perhaps: "When in doubt, err on the side of inclusion if it could prevent errors"?
…iles - Check for existing AGENTS.md in root directory - Check for existing AGENTS.md in all .roo/rules-*/ directories - Read and improve existing files rather than replacing them - Update discovery phase to explicitly look for these files
hannesrudolph
added
the
Issue/PR - Triage
…d files - Consolidate all AGENTS.md path checking into step 1 with CRITICAL emphasis - Add explicit instructions for deep dive improvement of existing files - Remove redundant mentions throughout the workflow - Emphasize iterating and enhancing existing files, not just appending - Add clear guidance on cleaning up outdated info and reorganizing - Make discovery phase the single source of truth for file checking
- Add explicit instructions to DELETE obvious information first - Emphasize that files should get SHORTER, not longer - Use stronger language: CRITICALLY EVALUATE, AGGRESSIVELY DELETE - Make clear that standard practices must be removed even if previously included - Add success metric: files should be more concise and valuable - Goal is to fix the tendency to just append instead of clean up
CRITICAL FIX: Paths were being interpreted as absolute, causing .roo to be created at system root - Clarify all paths are relative to PROJECT/WORKSPACE root - Add explicit warnings that .roo must be in project root, not system root - Update example to show proper project structure - Add IMPORTANT notes throughout emphasizing relative paths - Fix prevents accidental creation of /.roo at system level
- Replace 'mode-specific rule directories' with 'non-obvious' - Replace 'analysis_workflow' with 'discovered by reading files' - Tests now align with the new focus on non-obvious discoveries
Description
This PR updates the built-in
/initslash command to emphasize that it should only include non-obvious information discovered by reading files, making the generated AGENTS.md documentation more valuable and concise.Changes Made
1. Added CRITICAL Section
2. Updated Main File Output Structure
3. Transformed All Mode-Specific Examples
Code Mode now emphasizes:
safeWriteJson()requirement (prevents corruption)Debug Mode now emphasizes:
Ask Mode now emphasizes:
Architect Mode now emphasizes:
4. Updated Quality Criteria
Testing
Impact
This change will help AI assistants generate more valuable AGENTS.md files by focusing only on the information that truly matters - the non-obvious, project-specific knowledge that can't be guessed from standard practices.
Important
Refines
/initcommand to focus on non-obvious, project-specific information for AGENTS.md, excluding standard practices./initcommand inbuilt-in-commands.tsto focus on non-obvious, project-specific information for AGENTS.md.built-in-commands.spec.tsto check for non-obvious content in theinitcommand.initcommand content includes "non-obvious" and "discovered by reading files".This description was created by
for 8bcae24. You can customize this summary. It will automatically update as commits are pushed.