from my-cc-harness122
Performs a dual-pass code review of the current branch using Codex (GPT) and Claude, producing structured findings and severity classifications.
This skill automates a thorough code-review workflow: it runs a Codex-based first pass and a Claude second pass, aggregates findings into structured categories (CRITICAL/WARNING/INFO), and outputs a file-by-file report with suggested fixes. It includes commands to collect diffs, run static checks, and scan for secrets.
Use this skill when evaluating changes in the current branch or a pull request: functional correctness, security checks, performance patterns, tests coverage, and code-quality issues. It's intended for PR gating and developer feedback loops.
Designed for agents that can run shell commands and call LLM-based review tools (Codex/GPT-style and Claude). It expects Git access and local tooling (madge, gitleaks).
This skill has not been reviewed by our automated audit pipeline yet.
Harness Health Audit
Assess and score the overall health of a Claude Code harness across architecture, skills coverage, hooks, rules, MCP servers, eval pipelines, and team setup.
Code Evaluation Harness
Spawns an independent evaluator agent to score code outputs on functionality, code quality, originality, and usability/security, producing a structured EVAL_REP
SPEC-driven Development Interview
Conducts a structured, in-depth interview to produce a detailed SPEC.md requirements document for a feature or project.
Test-Driven Development (TDD)
Guides an agent through the TDD cycle: write a failing test, implement minimal code to pass, then refactor. Useful for disciplined feature development and maint
Plan
Create structured implementation plans for multi-step or architectural changes before coding — defines success criteria, trade-offs, and stepwise tasks.
