Move review methodology into agent system prompt - #7
Merged
Merged
Conversation
The review methodology now lives in agent/reviewer.md as the system prompt, making it deterministic — the agent cannot change or paraphrase it. command/review.md is slimmed to just pass the input. This prevents the spawning agent from taking liberties with the review prompt, which was breaking the ship cycle.
🤖 CodeAnt AI — Review Status
|
Thanks for using CodeAnt! 🎉We're free for open-source projects. if you're enjoying it, help us grow by sharing. Share on X · |
…prompt $ARGUMENTS is only substituted in command templates (prompt.ts:1383-1391). When placed in the agent system prompt (agent/reviewer.md), it's never replaced, so commands like `git show $ARGUMENTS` expand to empty strings in bash. Replaced with descriptive <input> placeholder; the concrete value is delivered via the user message (command/review.md: `Input: $ARGUMENTS`).
…r prompts Enforcement now lives in the plugin, not the prompt. Only two paths produce an enforcement-grade review note: 1. /review <target> command path (args.command === 'review') 2. Reviewer subtask with target-only prompt (SHA, PR ref, branch) Custom reviewer prompts (e.g. 'Review commit abc1234 for correctness...') still run but do NOT get a review note. This is the enforcement boundary. Prompt hardening in agent/reviewer.md is defense-in-depth only. Adds: - isCanonicalReviewInvocation() — pure function gating note attachment - isTargetOnlyPrompt() — rejects multiline prose, sentences, instructions - 13 new tests covering the exact bypass prompt from the live failure, /review path, target-only prompts, and rejection patterns
Adds note to Manual review section explaining that seeing only 'Input: <target>' is expected — the full methodology is in the reviewer agent system prompt. Updates PR description to document the canonical invocation gate and the user-visible behavior change.
CodeDeficient
added a commit
that referenced
this pull request
Jul 29, 2026
…argets extractPrNumber now matches both #7 and PR 7 / Review PR 7. isTargetOnlyPrompt rejects bare numeric strings like '7' before branch-name fallback, preventing ambiguous targets from resolving to HEAD. stripTargetSelectorPrefix no longer strips 'review pr ' prefix so PR 7 stays intact for extractPrNumber. Adds 8 new tests: - extractPrNumber: PR 7, Review PR 7, PR #7 - isCanonicalReviewInvocation: Review PR 7, PR 7 (true); bare '7', Review 7, Review branch 7 (false) - resolveTargetSha: PR 7, Review PR 7 resolve to PR head
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.
Why this exists
The review methodology lived in
command/review.mdas the command template. When the agent spawned a reviewer subagent viatask(), it was told to copy the prompt verbatim — but being an LLM, it paraphrased, truncated, or rewrote it. This broke the ship cycle: the reviewer subagent received an incomplete or wrong prompt.Design decisions
Decision: Move methodology into agent system prompt
The agent file body (
agent/reviewer.md) becomes the subagent's system prompt. This is deterministic — the spawning agent cannot change or paraphrase it. The command file is slimmed to just pass the input (Input: $ARGUMENTS).Alternatives considered:
tool.execute.beforehook to overwrite the prompt — more complex, more failure surfaceTradeoff: The methodology is now in the agent file, not the command file. Users who want a custom review prompt must edit
agent/reviewer.mdinstead ofcommand/review.md.Decision: Remove git notes mentions from agent-facing files
The plugin and pre-push hook are infrastructure. When agents know about git notes, they try to run
git notes addmanually, bypassing the plugin. The fix: remove allgit notesmentions from agent-facing instructions. The agent just needs to know/review <sha>before push.What this enables
Risks and safeguards
Risk: Users who customized
command/review.mdwill lose their custom promptMitigation: The methodology is now in
agent/reviewer.md— update that file insteadCodeAnt-AI Description
Make agent-initiated code reviews consistent and easier to start
What Changed
Impact
✅ Consistent review guidance✅ Fewer incomplete subagent reviews✅ Simpler agent-initiated reviews💡 Usage Guide
Checking Your Pull Request
Every time you make a pull request, our system automatically looks through it. We check for security issues, mistakes in how you're setting up your infrastructure, and common code problems. We do this to make sure your changes are solid and won't cause any trouble later.
Talking to CodeAnt AI
Got a question or need a hand with something in your pull request? You can easily get in touch with CodeAnt AI right here. Just type the following in a comment on your pull request, and replace "Your question here" with whatever you want to ask:
This lets you have a chat with CodeAnt AI about your pull request, making it easier to understand and improve your code.
Example
Preserve Org Learnings with CodeAnt
You can record team preferences so CodeAnt AI applies them in future reviews. Reply directly to the specific CodeAnt AI suggestion (in the same thread) and replace "Your feedback here" with your input:
This helps CodeAnt AI learn and adapt to your team's coding style and standards.
Example
Retrigger review
Ask CodeAnt AI to review the PR again, by typing:
Check Your Repository Health
To analyze the health of your code repository, visit our dashboard at https://app.codeant.ai. This tool helps you identify potential issues and areas for improvement in your codebase, ensuring your repository maintains high standards of code health.