MCP server exposing commit-check validation tools
Six tools with comprehensive descriptions and well-structured schemas. All tools have detailed parameter descriptions and clear output documentation. Tool names follow verb_noun pattern (validate_*). Strengths: excellent descriptions (150-300 chars each), complete parameter typing, clear error guidance in docstrings. Weaknesses: no enum constraints on config parameters; no explicit output schema definitions in code (only in docstrings); missing tool annotations in code despite being declared in features. The RESULT_SHAPE constant is well-designed but duplicated across all tool descriptions rather than composed. Parameter descriptions are thorough but could benefit from constraint examples (e.g., valid config keys).
Describe the commit-check validation rules enabled in a repository or by a config override. Use this to understand which rules are active before guessing at a format or suggesting a fix.
Validate author name and email against commit-check rules. Returns {status, warnings, checks[]}. status is 'pass', 'fail' or 'skip': only 'fail' is a rejection; 'skip' means every check skipped, so nothing was validated and the result is not approval. warnings is the number of checks with status 'warn'. Each check has: rule_id (stable rule id, e.g. 'CC001'); check (rule name, e.g. 'message'); status 'pass' | 'fail' | 'warn' | 'skip' ('warn' = the config lists the check under warn, so the finding is reported without failing the run; 'skip' = the rule did not run, e.g. the author is in ignore_authors or there was nothing to check); value (what was checked); error (why it failed); suggest (advice for a person); fix (the corrected value when the correction is unambiguous, else ''); docs_url (documentation for the rule). On 'fail', apply a non-empty fix verbatim; when fix is '', rewrite following suggest; then validate again.
Validate a branch name against commit-check rules. Returns {status, warnings, checks[]}. status is 'pass', 'fail' or 'skip': only 'fail' is a rejection; 'skip' means every check skipped, so nothing was validated and the result is not approval. warnings is the number of checks with status 'warn'. Each check has: rule_id (stable rule id, e.g. 'CC001'); check (rule name, e.g. 'message'); status 'pass' | 'fail' | 'warn' | 'skip' ('warn' = the config lists the check under warn, so the finding is reported without failing the run; 'skip' = the rule did not run, e.g. the author is in ignore_authors or there was nothing to check); value (what was checked); error (why it failed); suggest (advice for a person); fix (the corrected value when the correction is unambiguous, else ''); docs_url (documentation for the rule). On 'fail', apply a non-empty fix verbatim; when fix is '', rewrite following suggest; then validate again.
Config parameter lacks enum or constraint documentation. 'config' accepts arbitrary JSON objects with no guidance on valid keys (e.g., 'warn', 'commit.require_body'). LLMs may hallucinate invalid config keys.
Output schema documented only in docstrings (RESULT_SHAPE constant), not in code-visible JSON Schema. Clients cannot programmatically discover the structure of {status, warnings, checks[]} without parsing text.
Tool annotations (readOnlyHint, destructiveHint, idempotentHint, openWorldHint) are declared in ToolAnnotations objects but not visible in the tool registration. Verify they are correctly passed to mcp.tool() decorator.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 61 | 2026-07-28+ | v2 |
Validate a commit message, branch name, and author metadata together against commit-check rules. Returns {status, warnings, checks[]}. status is 'pass', 'fail' or 'skip': only 'fail' is a rejection; 'skip' means every check skipped, so nothing was validated and the result is not approval. warnings is the number of checks with status 'warn'. Each check has: rule_id (stable rule id, e.g. 'CC001'); check (rule name, e.g. 'message'); status 'pass' | 'fail' | 'warn' | 'skip' ('warn' = the config lists the check under warn, so the finding is reported without failing the run; 'skip' = the rule did not run, e.g. the author is in ignore_authors or there was nothing to check); value (what was checked); error (why it failed); suggest (advice for a person); fix (the corrected value when the correction is unambiguous, else ''); docs_url (documentation for the rule). On 'fail', apply a non-empty fix verbatim; when fix is '', rewrite following suggest; then validate again.
Validate a commit message against commit-check rules. Returns {status, warnings, checks[]}. status is 'pass', 'fail' or 'skip': only 'fail' is a rejection; 'skip' means every check skipped, so nothing was validated and the result is not approval. warnings is the number of checks with status 'warn'. Each check has: rule_id (stable rule id, e.g. 'CC001'); check (rule name, e.g. 'message'); status 'pass' | 'fail' | 'warn' | 'skip' ('warn' = the config lists the check under warn, so the finding is reported without failing the run; 'skip' = the rule did not run, e.g. the author is in ignore_authors or there was nothing to check); value (what was checked); error (why it failed); suggest (advice for a person); fix (the corrected value when the correction is unambiguous, else ''); docs_url (documentation for the rule). On 'fail', apply a non-empty fix verbatim; when fix is '', rewrite following suggest; then validate again.
Validate refs about to be pushed against commit-check rules, including force-push safety. Returns {status, warnings, checks[]}. status is 'pass', 'fail' or 'skip': only 'fail' is a rejection; 'skip' means every check skipped, so nothing was validated and the result is not approval. warnings is the number of checks with status 'warn'. Each check has: rule_id (stable rule id, e.g. 'CC001'); check (rule name, e.g. 'message'); status 'pass' | 'fail' | 'warn' | 'skip' ('warn' = the config lists the check under warn, so the finding is reported without failing the run; 'skip' = the rule did not run, e.g. the author is in ignore_authors or there was nothing to check); value (what was checked); error (why it failed); suggest (advice for a person); fix (the corrected value when the correction is unambiguous, else ''); docs_url (documentation for the rule). On 'fail', apply a non-empty fix verbatim; when fix is '', rewrite following suggest; then validate again.
Parameter descriptions reference example config structures (e.g., '{"warn": ["message"]}') which LLMs may reuse literally. Replace with formal constraint documentation or link to config schema.
describe_validation_rules lacks guidance on when to call it. Description should state: 'Call this before validate_* to understand which rules are active and avoid invalid suggestions.'