Skill Claude CodeCodex
Improve AI application with evaluation-driven development. Define eval criteria, instrument the application, build golden datasets, observe and evaluate application runs, analyze results, and produce a concrete action plan for improvements. ALWAYS USE THIS SKILL when the user asks to set up QA, add tests, add evals…
Skill Claude CodeCodex
Use when designing evaluation suites for agentic systems, choosing agent benchmarks, measuring multi-agent performance, or critiquing single-score evaluation of agents. Based on arXiv:2605.26112v1. Trigger phrases: "evaluate my agent", "benchmark agent", "measure agent reliability", "pass^k evaluation", "agent metrics…
Skill Claude CodeCodex
Set up and optimize a complete local AI infrastructure — Ollama, Open WebUI, local RAG pipelines, and private model serving. Use when running models locally for privacy, compliance, offline use, or cost, or setting up GPU-accelerated local inference and RAG.
Skill Claude CodeCodex
Use when evaluating MCP-enabled agents, tool-using LLM systems, multi-server workflows, tool-selection quality, trajectory quality, task completion, or MCP security regressions.
Skill Claude CodeCodex
Merges mature lessons from a domain memory file into its instruction file. Syntax: /memory-merger >domain [scope] where scope is global (default), user, workspace, or ws. Use when a domain memory file has accumulated lessons that should become permanent instructions.
Skill Claude CodeCodex
Generate or edit images via OpenRouter with the Gemini 3 Pro Image model. Use for prompt-only image generation, image edits, and multi-image compositing; supports 1K/2K/4K output.
Skill Claude CodeCodex
Use when comparing basic, enhanced, GraphRAG, or agentic RAG designs, evaluating domain-specific RAG quality, tuning retrieval components, or deciding whether agentic RAG is worth its cost.
Skill Claude CodeCodex
Create, update, refactor, explain, or review Semantic Kernel solutions using shared guidance plus language-specific references for .NET and Python. Use when working on applications, plugins, or function-calling flows built on Semantic Kernel.
Skill Claude CodeCodex
Use when improving, evolving, benchmarking, or maintaining agent skills over time. Treats skills as long-lived structured artifacts: execute tasks, diagnose trajectory evidence, accumulate textual momentum, patch at the correct abstraction layer, test per model and harness, canary changes, and roll back regressions.
Skill Claude CodeCodex
Creates ASP.NET Minimal API endpoints with correct request/response types and complete OpenAPI (Swagger) documentation. Use when adding or reviewing Minimal API endpoints in .NET, or when generated OpenAPI output is missing schemas, status codes, or descriptions.
Skill Claude CodeCodex
Use when the user asks to work on a Python FastAPI project. Triggers on keywords Python, FastAPI, Pydantic, SQLAlchemy.
Skill Claude CodeCodex
Apply GDPR-compliant engineering practices across your codebase. Use this skill whenever you are designing APIs, writing data models, building authentication flows, implementing logging, handling user data, writing retention/deletion jobs, designing cloud infrastructure, or reviewing pull requests for privacy…
Skill Claude CodeCodex
Use when a GraphQL @defer/@stream multipart response parses but a patch disappears after client, proxy, cache, or SSR hydration. Replay a pinned current ID-based or legacy path-based envelope offline, reject unknown shapes, and make a bounded rollout or recovery decision.
Skill Claude CodeCodex
Use when gRPC status or trailing metadata disappears or changes across proxies, gateways, meshes, tunnels, or runtimes, especially for trailers-only and size-boundary failures.
Skill Claude CodeCodex
Use when an HTTP download gets the wrong, garbled, unsafe, or client-dependent filename—especially Content-Disposition filename/filename, RFC 8187 percent encoding, duplicate parameters, redirects, path traversal, device names, or media-type extension mismatches.
Skill Claude CodeCodex
Use when HTTP 103 Early Hints may be dropped, merged, reordered, or corrupted across reverse proxies—replay bounded synthetic exchanges, compare normalized hop captures, and preserve the final response.
Skill Claude CodeCodex
Use when an HTTP client follows redirects and credentials are silently dropped or may cross host, scheme, or port boundaries—especially Authorization, Cookie, Proxy-Authorization, API-key headers, 301/302/303/307/308, or default-port confusion.
Skill Claude CodeCodex
Use when scoped or link-local IPv6 input containing an interface name/index works in one CLI, URL parser, HTTP client, resolver, or OS but fails or changes in another. Classify UI, URI, socket, and wire boundaries offline before any network probe.
Skill Claude CodeCodex
Use when JSON authorization, signing, canonicalization, APIs, or distributed runtimes may disagree on repeated object names—inspect original bytes before lossy parsing and fail closed with scoped duplicate paths and offsets.
Skill Claude CodeCodex
Use when multipart/form-data uploads succeed in one emitter/parser but fail, truncate, or disagree through another runtime or proxy. Inspect redacted raw bytes offline, reject ambiguous framing and disposition parameters, compare normalized part trees, and make a bounded rollout decision.
Skill Claude CodeCodex
Use when OAuth DPoP clients, authorization servers, or resource servers disagree on DPoP-Nonce challenges, proactive rotation, endpoint scope, CORS exposure, concurrency, or bounded retry behavior.
Skill Claude CodeCodex
Use when OpenAPI 3.0 nullable schemas are ignored, rewritten, or generate unusable types after an OpenAPI 3.1 migration. Inventory requiredness, null acceptance, references, composition, and fixtures offline before changing a contract.
Skill Claude CodeCodex
Generates a complete, working application from an OpenAPI specification using the active framework's conventions: models, routes, validation, and tests. Use when scaffolding a service from an existing OpenAPI/Swagger spec rather than hand-writing endpoints.
Skill Claude CodeCodex
Also use for async Python patterns, Pydantic schemas for LLM structured outputs, retry strategies for LLM calls, background AI jobs, or any service combining FastAPI and LLMs. If Python + AI backend is in scope — use this skill.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: