Comprehension Gate — Development agent for Claude Code
Cold-reads code review comment bodies with no diff and no code access and judges whether a reader can tell what breaks and what to do from the body alone.
How to install Comprehension Gate
Installs to ~/.claude/agents/haacked-review-code-comprehension-gate.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/haacked/review-code/HEAD/agents/comprehension-gate.md -o ~/.claude/agents/haacked-review-code-comprehension-gate.md Restart Claude Code, or start a new session, for it to be picked up.
What Comprehension Gate does
name: comprehension-gate description: "Cold-reads code review comment bodies with no diff and no code access and judges whether a reader can tell what breaks and what to do from the body alone. Returns PASS or REWRITE per item with a note on what was unclear. Never rewrites text and never judges technical correctness. Use after finding validation and the adversary meta-review, before the voice pass." model: haiku color: cyan metadata: execution-tier: fast
**Your entire response is a s
Alternatives in Development
- Get Commit ID — gh api repos/OWNER/REPO/pulls/PR_NUMBER --jq '.head.sha' gh api repos/OWNER/REPO/pulls/PR_NUMBER/comments \ 1.3k ★
- Clean Code Refactor — Use this agent when you need to refactor existing code, rewrite a module for better maintainability, decompose 358 ★
- Quality Verifier — 웹소설 에피소드 품질 검증 전문가 144 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Code Reviewer Voice
Rewrites code review comment bodies in plain, conversational voice. Preserves every citation, file path, line
PR Comment Cleaner
This agent should be invoked to remove and rewrite code comments in the files a pull request changes. This inc
Invest Gate
Returns PASS/FAIL per INVEST letter for a given issue body, with one-line reasoning per criterion. Invoke when
Sb Distiller
Rewrites strictlybetter's inheritance body from the ledger for a cold-start reader. MUST BE USED by /strictlyb
Docplan Auditor
Cold semantic-grouping judgment for a DocPlan. Given ONLY the DocPlan (its sections + excluded[]) and the Conv
Comment Accuracy
Detects stale comments, docstring-signature mismatches, and contradicted NOTE/TODO/FIXME markers
Related Skills
Reconstruction And Visualisation Skill
Claude skill to create an output optimised for the reader's comprehension from existing source material (a mes
Rewrite History
Reimplement the current branch with a clean, narrative-quality git commit history suitable for reviewer compre
Opencode Plugin Comment Judge
opencode plugin: a model judges every comment your agent writes and removes or rewrites the ones that do not e