fix: improve agent-native-reviewer with triage, prioritization, and stack-aware search#387
Merged
Merged
Conversation
…tack-aware search - Add missing frontmatter fields (color, tools) to match peer review agents - Add triage step (step 0) to orient the agent in unfamiliar codebases - Add stack-specific search strategies for common frameworks (Vercel AI SDK, LangChain, OpenAI Assistants, Claude Code plugins, Rails + MCP) - Add prioritization heuristics so not every missing tool is a Critical finding - Add incremental vs. full audit guidance for the most common review scenario - Add confidence calibration to align with the ce-review ensemble pattern - Add "What You Don't Flag" section to prevent false positives on intentionally human-only flows (CAPTCHA, 2FA, OAuth consent, biometrics) - Restore the "Noun Test" heuristic as step 6 in the review process - Nuance "primitives over workflows" with justified exceptions - Collapse verbose anti-pattern subsections into a scannable reference table - Cut from 262 lines to 192 while adding substantial new guidance
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 2a42f5111a
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
- Align noun-test severity with priority tiers from step 2
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.
The agent-native-reviewer had the right philosophy but couldn't navigate real codebases -- it told the agent WHAT to check without helping it figure out HOW to find things in an unfamiliar project. This rewrites the system prompt to close that gap.
Key changes:
Also adds missing frontmatter fields (
color: cyan,tools: Read, Grep, Glob, Bash) and collapses the original's 7 verbose anti-pattern subsections into a scannable reference table. Net result: 192 lines down from 262, with substantially more actionable guidance.🤖 Generated with Claude Opus 4.6 (1M context, extended thinking) via Claude Code