Skill Claude CodeCodex
Triage a GitHub issue by verifying all factual claims against running code, assessing roadmap fit, then posting a structured response comment with confirmed findings, stale corrections, a recommendation, and concrete test steps.
Skill Claude CodeCodex
Use when writing, reviewing, or refactoring Go code — naming conventions, receiver naming, error wrapping, interface design, goroutine safety, and common pitfalls (goroutine leaks, defer-in-loop, nil map writes).
Skill Claude CodeCodex
Triggers on harness friction events (hook failures, ambiguous guidance, violated mandates, mismatched reality) encountered during any session. Agent must: address the immediate problem, log it locally to docs/operational/harness-feedback.md, and file it upstream to andr-ca/agentharness. Default-on; skip upstream…
Skill Claude CodeCodex
Use when adding logging to an application, reviewing log output, choosing log levels, structuring logs for observability, or configuring logging backends — covers structured logging, YAML config patterns, what NOT to log, and local vs. production output.
Skill Claude CodeCodex
Use when two or more agent sessions may work on the same repository concurrently — covers the per-feature lock-file protocol, stale-lock detection, worktree isolation rules, and what to do when a feature is already locked.
Skill Claude CodeCodex
Use when writing tests for critical business logic, auditing test suite quality beyond line coverage, or interpreting surviving mutants — covers mutation operators, mutation score thresholds, and tooling (mutmut, Stryker, gremlins) for Python, TypeScript/JS, and Go.
Skill Claude CodeCodex
Use when diagnosing slow code, high memory usage, or CPU spikes — language-agnostic profiling workflow: form a hypothesis, identify the hot path, benchmark before and after, and interpret profiler output. Includes Python (cProfile, py-spy), Go (pprof), and Node.js (--inspect, clinic.js) tooling.
Skill Claude CodeCodex
Use when starting a complex multi-step task, a research project, or any task requiring more than 5 tool calls — creates and maintains taskplan.md, findings.md, and progress.md to preserve state across context resets.
Skill Claude CodeCodex
Use when asked to port, migrate, or add equivalent agent instructions, skills, or custom sub-agents for a different coding tool than the one already configured — e.g. "add Cursor support from our CLAUDE.md", "we're switching from Cursor to Codex, port our rules", "make this work for Copilot too", "port our review…
Skill Claude CodeCodex
Use on the first run in a project, or when asked to set the harness up, tailor it to this repository, or work out which conventions apply here — inventories what the project already does, interviews the owner about the decisions only they can make, and applies a confirmed plan without overwriting existing setup.
Skill Claude CodeCodex
Use when writing or reviewing Python code — naming conventions, type hints, common pitfalls (mutable defaults, bare except, is-vs-==), and testing structure.
Skill Claude CodeCodex
Use when writing, reviewing, or optimizing React or Next.js code — component architecture, hooks rules, server vs. client components, data fetching patterns, bundle size, and accessibility in React.
Skill Claude CodeCodex
Use before implementing a significant feature or change when requirements are ambiguous, underspecified, or likely to have multiple reasonable interpretations, or when a requirement arrives pre-formed as an issue, audit finding, or review comment whose claim needs checking — covers verifying the premise before acting…
Skill Claude CodeCodex
Use when reviewing code for security vulnerabilities, performing a security audit, or checking for OWASP Top 10 issues — covers injection flaws, broken access control, cryptographic failures, secrets exposure, dependency vulnerabilities, and language-specific pitfalls for Python, TypeScript/JavaScript, and Go.
Skill Claude CodeCodex
Use when designing classes or APIs to evaluate SOLID compliance — SRP, OCP, LSP, ISP, DIP. Covers what each principle means, how to diagnose violations by symptom, and how to refactor.
Skill Claude CodeCodex
Use when writing tests, deciding on test coverage, choosing between unit/integration/E2E tests, applying TDD, or reviewing test quality — covers rigor tiers, the Red-Green-Refactor cycle, coverage requirements, and Playwright for UI testing.
Skill Claude CodeCodex
Use when writing, reviewing, or refactoring TypeScript or JavaScript code — naming conventions, type annotations, private fields, null vs. undefined, async/await pitfalls, and module structure.
Cursor rule Cursor
Use when building or reviewing web UIs for accessibility — WCAG 2.2 Level AA compliance, semantic HTML over ARIA, keyboard navigation, color contrast, screen reader patterns, and testing approach.
Cursor rule Cursor
Generated from this repo's own CLAUDE.md by tools/generate-cursor-rules.sh (https://github.com/andr-ca/agentharness) — do not hand-edit; regenerate instead (tools/generate-cursor-rules.sh --output-dir .). A CI check keeps this file in sync with its source (see .github/workflows/ci.yml's content-quality job).
Cursor rule Cursor
Use when building multi-turn agents, tool-calling systems, agent orchestration, or autonomous workflows — covers loops, tool calling, branching, reflection patterns.
Cursor rule Cursor
Use when designing, reviewing, or evolving a REST or GraphQL API — resource naming, HTTP status codes, versioning strategy, error response shapes, pagination, and authentication patterns.
Cursor rule Cursor
Use when asked to check whether review/audit recommendations were actually implemented, whether gaps were closed, or to re-score the repo — verifies claims against repo state instead of trusting status reports.
Cursor rule Cursor
Use before implementing big work — new systems, pipelines, daemons, or anything stateful/concurrent — where a bug would be expensive to find after code exists. Requires a written implementation plan to pass multi-round adversarial review by two independent LLM reviewers before any code is written. Not for scoped bug…
Cursor rule Cursor
Use when creating a branch, naming it, deciding whether to use a worktree, or handling secrets accidentally committed to history — branch naming convention, trunk protection, and secrets-removal procedure.