Claude Skills by rakib-nyc

Audits AI coding-agent context. Measures always-on context-window cost of CLAUDE.md, AGENTS.md and Agent Skills, and reports skills that will not load, stale path references, redundant instructions and duplicate skill names. Use when the user asks to audit, measure or reduce context cost, check whether a skill is valid or will load, review CLAUDE.md or AGENTS.md quality, or find duplicate, conflicting or unused skills.

Valid skill preceded by a UTF-8 byte order mark. Use when testing BOM handling.

Valid skill using Windows line endings. Use when testing CRLF handling.

body

Frontmatter name is a number, not a string. Use when testing type validation.

"broken

body

Frontmatter omits name; the harness derives it from the directory. Use when testing name inference.

Just prose, no frontmatter fence at all.

Explains tokenizer control sequences such as <|endoftext|> and <|im_start|>. Use when the user asks about tokenizer internals.

Skill with emoji 🎯, CJK 日本語テキスト, RTL نص عربي, and combining marks é. Use when testing unicode.

--- name: never-closed description: The fence is never closed. Use when testing. body continues forever

Reference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claude/Anthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`, `[1m]`); the user asks about an LLM (pricing/model choice/limits/caching) — never answ...

Creates accessible React components for the internal admin console. Use when the user asks to add or modify a screen, form, dialog, or table in the admin app.

Builds parameterised SQL against the ledger schema. Use when the user asks to read from or write to the accounts, postings, or journals tables.

Drafts customer-facing release notes from merged pull requests. Use when the user asks to prepare a changelog or announce a version.

Builds new web components. Use when the user asks to write or restyle a React component.

Applies house styling rules to web code. Use when the user asks to write or restyle a React component.

Organises Python import blocks. Use when the user asks to sort or clean up imports in a .py module.

Formats Go source files. Use when the user asks to reformat, lint, or apply style rules to a .go file.

Formats Python source files. Use when the user asks to reformat, lint, or apply style rules to a .py file.

Reviews Python code for performance defects. Use when the user asks to profile hot paths, reduce allocations, remove N+1 database round trips, or speed up a slow endpoint.

Reviews Python code for security defects. Use when the user asks to check code for injection, authentication bypass, unsafe deserialization, or secret leakage.

Performs a code review pass. Use when the user asks to review code before merging.

Reviews a diff for defects. Use when the user asks to review a pull request or inspect changes.

Generates a new agent skill from a template. Use when the user asks to build a skill or add agent capabilities.

Scaffolds a new agent skill directory. Use when the user asks to create a skill or add a new capability.

Deploys a service to production. Use when the user asks to ship, release, or roll back.

Deploys a service to production. Use when the user asks to ship, release, or roll back.

Deploys a service to production. Use when the user asks to ship, release, or roll back.

Runs the repo linter. Use when the user asks to lint or format the monorepo.