Skills for AI coding agents

41 skills

A skill is a SKILL.md file that teaches Claude Code, Codex or Cursor one job: review a pull request, find the root cause of a bug, write a safe migration. Pick one, read what it tells the agent, and install it with one command.

Run these skills on any model

Skills work with whatever model your agent is pointed at. Connect Claude Code, Codex or Cursor to AnyModel and run them on Claude, GPT, Gemini and more with one key, billed per token.

PR review

98 lines

Reviews a pull request, a branch or uncommitted changes, putting correctness bugs first and design second, with severity-ranked findings that cite file and line. Use when the user asks to review a PR, a branch or a diff.

Code reviewCode Review

Security review

80 lines

Audits a code change for injection, broken authorization, leaked secrets, SSRF, path traversal, unsafe deserialization, weak crypto and missing rate limits by tracing untrusted input to sinks. Use when the user asks for a security review or audit of a change.

SecurityCode Review

Performance review

75 lines

Reviews code for performance problems such as N+1 queries, work inside hot loops, unbounded memory growth, missing indexes, blocking I/O on the request path and chatty network calls. Use when the user asks why code is slow or to review a change for performance.

Code reviewCode Review

Responding to review feedback

89 lines

Works through reviewer comments on a pull request by verifying each claim against the code, fixing what is right, pushing back with evidence where it is wrong, and drafting a reply per comment. Use when the user shares review comments to address.

Code reviewCode Review

Self-review before PR

97 lines

Runs a checklist pass over your own diff before opening a pull request, catching debug leftovers, dead code, secrets, missing tests, poor naming, scope creep and migration or config impact. Use when the user is about to open a PR or says the work is done.

Code reviewCode Review

Root-cause debugging

82 lines

Debugs a bug by reproducing it first, isolating it, testing one hypothesis at a time with evidence, fixing the cause instead of the symptom and adding a regression test. Use when something fails, crashes or behaves unexpectedly.

DebuggingDebugging

Log triage

100 lines

Reads logs and stack traces to find the first real failure, correlates events by request id and timestamp, separates the cause from cascade noise, with grep and jq recipes. Use when asked to investigate errors in logs, an incident or a crash.

DebuggingDebugging

Git bisect for regressions

101 lines

Finds the commit that introduced a regression using git bisect, manually or automated with git bisect run and a script, and handles flaky or unbuildable commits with skip. Use when something worked before and now fails.

DebuggingDebugging

Memory leak hunt

103 lines

Confirms a memory leak under steady load, then locates it with heap snapshots and diffs in Node.js, Python tracemalloc or a JVM heap dump, and checks common sources such as unbounded caches and listeners. Use when memory grows over time or a process gets OOM-killed.

DebuggingDebugging

Flaky test hunter

82 lines

Finds why a test passes and fails intermittently by looping runs, shuffling order and checking timing, shared state, time zones, network and unordered collections, then fixes the cause. Use when CI is flaky or a test fails only sometimes.

TestingDebugging

Test-driven development

91 lines

Drives a change with the red, green, refactor loop: write a failing test that states the behavior, add the minimal code to pass, then refactor with tests green. Use when building a feature or fixing a bug test-first, or when asked for TDD.

TestingTesting

Unit test writer

74 lines

Writes focused unit tests for existing code by finding its public behavior, covering edge cases and errors, using meaningful assertions and following the project's test framework and style. Use when asked to add tests for a function, module or class.

TestingTesting

E2E tests with Playwright

103 lines

Writes stable browser end-to-end tests with Playwright using user-facing locators, web-first assertions instead of sleeps, isolated test data, traces on failure and headed debugging. Use when asked to add or fix browser E2E tests.

TestingTesting

Test coverage gaps

91 lines

Finds risky untested code by running coverage and ranking gaps by risk such as money, auth, data loss and parsing rather than by percentage, then proposes the few tests that matter most. Use when asked where tests are missing or to improve coverage.

TestingTesting

API contract tests

82 lines

Writes HTTP API tests that check status codes, response schemas, error shapes, auth and permission cases, idempotency and pagination while staying independent of database state. Use when asked to test a REST or JSON API.

TestingTesting

Dockerfile review

101 lines

Writes and reviews Dockerfiles for small, reproducible, secure images: pinned bases, multi-stage builds, cache-friendly layer order, non-root user, .dockerignore. Use when the user creates, edits or asks to review a Dockerfile or container image.

DevOpsDevOps

GitHub Actions CI

111 lines

Writes and fixes GitHub Actions workflows: triggers, caching, matrix builds, concurrency, least-privilege permissions, secrets, SHA-pinned actions, and log-based debugging. Use when the user adds, edits or debugs a CI workflow or a failing GitHub Actions run.

DevOpsDevOps

Kubernetes manifest review

85 lines

Reviews Kubernetes manifests and Helm values for reliability and security: resources, probes, securityContext, image tags, PodDisruptionBudget, rollout strategy. Use when the user writes, edits or asks to review K8s YAML, a Helm chart or a deployment.

DevOpsDevOps

Incident runbook

106 lines

Guides triage of a production incident: assess impact, stabilize first by rollback or feature flag, keep a timeline, communicate, then write a blameless postmortem. Use when the user reports an outage, errors in production, a failed deploy or asks for a postmortem.

DevOpsDevOps

Release checklist

97 lines

Runs a pre-release and post-release checklist: migrations, config and env changes, feature flags, backward compatibility, monitoring, rollback plan and release notes. Use when the user is about to ship, deploy or tag a release and wants a go/no-go review.

DevOpsDevOps

Env and config audit

86 lines

Finds hardcoded secrets and config in code, moves them to environment variables, maintains a documented .env.example, ensures .env is git-ignored and checks git history for leaks. Use when the user asks to audit secrets, clean up config or before publishing a repo.

SecurityDevOps

Commit message writer

97 lines

Writes commit messages from the staged diff with a short imperative subject and a body that explains why, one logical change per commit, matching the repo's existing convention. Use when the user asks to commit, or for a commit message for staged changes.

Git & PRsGit Workflow

PR description writer

96 lines

Writes a pull request description from the diff and commits covering what changed and why, how it was tested, risks and rollout, and screenshots for UI, without inventing claims. Use when the user asks for a PR description or is about to open a PR.

Git & PRsGit Workflow

Merge conflict resolver

94 lines

Resolves git merge and rebase conflicts by reading the intent of both sides and their commits, never blindly taking ours or theirs, then rebuilding and testing, and asking the user when intents truly clash. Use when a merge or rebase stops on conflicts.

Git & PRsGit Workflow

Changelog writer

91 lines

Produces a user-facing changelog or release notes from the commits and PRs since the last tag, grouped into Added, Changed, Fixed and Removed, skipping internal noise. Use when the user asks for release notes, a changelog entry or what changed since the last release.

DocumentationGit Workflow

Safe git recovery

104 lines

Recovers from git mistakes without losing work using reflog, revert versus reset, restoring deleted branches, undoing bad merges or rebases and finding lost stashes, backing up first and asking before anything destructive. Use when the user broke or lost something in git.

Git & PRsGit Workflow

SQL query review

94 lines

Reviews SQL queries and ORM code for correctness (NULLs, duplicating joins, time zones) and performance (EXPLAIN, indexes, SELECT *, N+1, keyset pagination) with parameterized queries. Use when the user writes, debugs or optimizes SQL or a slow query.

Databases & SQLDatabases

Safe DB migration

96 lines

Plans and writes zero-downtime schema migrations: expand/contract, batched backfills, concurrent indexes, online DDL, safe defaults, lock timeouts and rollback. Use when the user adds or changes columns, indexes or constraints on a database used in production.

Databases & SQLDatabases

Spec-first kickoff

81 lines

Turns a vague app idea into a one-page spec and a step-by-step build plan before any code is written, asking only the questions that change the design. Use when the user describes a new project, feature or app idea and wants to start building.

Vibe codingVibe Coding

Small steps builder

68 lines

Builds software in small verifiable increments - one change, run it, commit, repeat - keeping the app runnable at every step and recovering cleanly when a step breaks things. Use when implementing a feature or app through a multi-step build with an agent.

Vibe codingVibe Coding

Codebase explainer

73 lines

Maps an unfamiliar repository quickly and produces a short orientation doc with entry points, build, run and test commands, directory roles, request and data flow, key abstractions and gotchas, with file references. Use when the user is new to a repo or asks how a codebase works.

CodingVibe Coding

Prototype to production

69 lines

Hardens a working prototype with a prioritized list of fixes instead of a rewrite - error handling, input validation, env config, logging, core-flow tests, auth checks, pinned dependencies and deploy notes. Use when a demo or vibe-coded app needs to go live.

Vibe codingVibe Coding

UI polish pass

73 lines

Fixes common problems in AI-built interfaces - inconsistent spacing, missing empty, loading and error states, broken phone layouts, weak focus and contrast, text overflow, mismatched buttons - and verifies with screenshots at 390px and 1440px. Use when a UI works but looks rough.

Vibe codingVibe Coding

Safe refactoring

76 lines

Refactors code without changing behavior - lock current behavior with tests first, apply small mechanical steps, run tests after each, keep refactor commits separate from behavior changes, and stop when scope grows. Use when asked to clean up, restructure, rename or simplify code.

CodingVibe Coding

README writer

70 lines

Writes or fixes a README from the actual code - what the project is, a quick start whose commands are verified, a configuration table, common tasks and troubleshooting - without inventing features. Use when a repo has no README, an outdated one, or the user asks for project docs.

DocumentationDocumentation

API docs writer

84 lines

Documents HTTP endpoints from the real code - method, path, auth, parameters, request and response examples, error codes - so examples match the actual handlers. Use when a service has undocumented or outdated API docs, or the user asks to document an API.

DocumentationDocumentation

Architecture decision record

71 lines

Writes a short, dated architecture decision record (ADR) with context, decision, options considered with trade-offs, consequences and status, stored next to the code. Use when a significant technical choice was made or is being made and should be recorded.

DocumentationDocumentation

Code comments cleanup

78 lines

Cleans up code comments - removes noise that restates the code, keeps and sharpens why-comments, fixes stale comments that lie, and adds a comment only where intent is non-obvious. Use when asked to tidy, review or improve comments in a file, module or diff.

DocumentationDocumentation

AnyModel API

125 lines

Calls Claude, GPT, Gemini and other models through the AnyModel OpenAI- and Anthropic-compatible API with one key, and wires existing code or agents to it. Use when the user wants to call an LLM from code, switch providers, or connect a tool to AnyModel.

AI models & APIsAnyModel

Model picker

83 lines

Picks the right LLM for a task by weighing capability, context window, latency and price, and backs the choice with a quick side-by-side test on real inputs. Use when the user asks which model to use, wants a cheaper model, or a model fails at a task.

AI models & APIsAnyModel

LLM cost optimizer

77 lines

Finds where an application spends money on LLM calls and cuts it with model routing, prompt trimming, caching, batching and output limits, measuring cost and quality before and after. Use when the user says the AI bill is too high or wants to make LLM usage cheaper.

AI models & APIsAnyModel

Skills in this catalog are written by the AnyModel team and licensed under CC BY 4.0. Copy, use and change them, crediting AnyModel with a link. A skill tells your agent what to do, so read it before you install it. For community plugins and skill packs, see Plugins & skills.