Agent skill
oh-my-ccg-review
独立双模型交叉审查(随时可用)
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/oh-my-ccg-review
SKILL.md
oh-my-ccg Review Phase
You are executing an independent dual-model cross-review. This can be run at any time.
Workflow
Step 1: Collect Artifacts
Determine what to review:
- If RPI state exists with a change: use
rpi_state_readMCP tool to load the change's implementation details - If git diff has changes: review the diff
- Otherwise: ask user what to review
Step 2: Dual-Model Review (PARALLEL — MANDATORY)
Launch BOTH models in a single message using MCP tools:
Codex (code-reviewer role) — use ask_codex tool:
agent_role: "code-reviewer"prompt: Include the code to review, ask for logic/security analysiscontext_files: All changed filesbackground: true
Gemini (designer role) — use ask_gemini tool:
agent_role: "designer"prompt: Include the code to review, ask for pattern/maintainability analysisfiles: All changed filesbackground: true
Then use check_job_status for both job IDs to collect results.
Step 3: Synthesize Findings
Merge findings from both models:
- Deduplicate overlapping issues
- Classify severity: Critical / Warning / Info
- Cross-reference: issues found by BOTH models get elevated severity
Step 4: Present Report
## Review Report
### Critical (must fix)
- [C1] file:line — description (found by: Codex/Gemini/Both)
### Warning (should fix)
- [W1] file:line — description
### Info (nice to have)
- [I1] file:line — description
### Summary
- Total findings: X (C: N, W: N, I: N)
- Cross-validated: N issues found by both models
Step 5: Decision Gate
Ask user:
- Fix: Address Critical/Warning issues → loop back to executor
- Accept: Archive the review, proceed
- Defer: Note issues for later, proceed
If fixing: use rpi_state_write MCP tool to record review findings and transition back to impl.
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?