Install interactively
npx skills@latest add kuraudo-lab/skillsagent skills catalog
Skills
A personal catalog of reusable agent workflows from kuraudo-lab/skills , organized by category and installable with the open skills CLI.
// Install
Install Instructions
Install globally for Codex
npx skills@latest add kuraudo-lab/skills --global --agent codexList available skills
npx @kuraudo-lab/skills listInstall one skill
npx skills@latest add kuraudo-lab/skills --skill create-git-commit --global --agent codex// Catalog
Skills by Category
14 skills synced from GitHub.
Development
8create-git-commit
Generate a Conventional Commits message from staged Git changes and create the commit. Use when the user asks to commit changes, make a commit, generate a commit message, use Conventional Commits, or review staged changes before committing.
phoenix-audit-deletion-readiness
Audit whether a module, feature, service, or code path can be safely deleted and regenerated from durable intent and evaluations. Use when the user asks for deletion readiness, regeneration safety, Phoenix deletion test, whether code can be removed, module replacement risk, or survivability of specs and tests after deleting implementation.
phoenix-build-evaluation-registry
Build or update a Phoenix-style evaluation registry that maps requirements, invariants, and acceptance criteria to tests, checks, reviews, metrics, and live evidence. Use when the user asks to catalog evaluations, find durable tests, map specs to tests, create an eval registry, classify ephemeral/durable/live evaluations, or prepare deletion/regeneration safety evidence.
phoenix-canonicalize-spec
Extract canonical requirements, constraints, decisions, invariants, and acceptance criteria from specs, design docs, tickets, or conversation context. Use when the user asks to canonicalize a spec, structure requirements, create Phoenix clauses, normalize intent, split a design into stable clauses, or prepare inputs for an intent graph.
phoenix-capture-implementation-memory
Extract hidden implementation knowledge from existing code before refactoring, replacing, deleting, or regenerating it. Use when the user asks to preserve scar tissue, understand legacy logic, capture compatibility behavior, avoid losing production lessons, inspect special cases, or prepare a Phoenix implementation memory record.
phoenix-create-provenance-commit
Record a design or intent provenance artifact and create a Git commit before implementation. Use when the user asks for provenance, conversation commit, design commit, intent trail, decision record, Phoenix Architecture provenance, or wants design changes and agent conversation context committed before coding.
phoenix-link-implementation-provenance
Link implementation work to an existing provenance record and ensure implementation commits include provenance trailers. Use when coding from a provenance commit, checking whether implementation references design intent, preparing an implementation commit after phoenix-create-provenance-commit, or enforcing lightweight Phoenix provenance links.
phoenix-link-intent-graph
Maintain a lightweight file-backed intent graph connecting requirements, constraints, decisions, provenance, evaluations, implementation files, commits, and evidence. Use when the user asks to link intent, build an intent graph, map requirement-to-test-to-code relationships, update Phoenix graph files, trace why code exists, or prepare selective invalidation analysis.
Knowledge
1web-knowledge-note
Turn a source webpage into a structured Obsidian knowledge note with source links, concise summaries, and important diagrams/images preserved. Use when the user asks to整理/总结/梳理 a webpage, article, documentation page, or online guide into a target folder or Markdown note, especially when they mention keeping key illustrations.
Learning
1generate-exam
Generate a self-contained, self-grading static HTML exam from a user-specified scope of local documents. Use when the user asks to generate an exam, create an exam, quiz, knowledge check, mock test, practice paper, or assessment from Markdown, text, code, or other readable workspace files, using objective questions or fill-in-the-blank questions with exact answers.
Research
1digest-product-hunt-ideas
Turn Product Hunt launches and comments into Mac/iOS app ideas, including trend extraction, native-platform opportunity analysis, MVP scoping, and scoring. Use when scanning Product Hunt, hunting indie app ideas, evaluating Mac/iOS product opportunities, or generating idea radar reports.
Workflow
2curate-personal-skills
Turn accumulated personal workflows, prompts, notes, scripts, and project conventions into concise reusable agent skills. Use when organizing a personal skills repository, extracting a skill from repeated work, reviewing skill quality, or preparing skills for installation through the open skills CLI.
maintain-skills-package
Maintain an installable agent skills repository backed by the open skills CLI. Use when adding, validating, listing, publishing, or troubleshooting a skills package that should work with commands like npx skills@latest add owner/repo.
Writing
1post-writer
Draft and revise platform-optimized posts for Medium, WeChat Official Accounts, and Xiaohongshu from user-provided key points. Use when the user asks to write, polish, adapt, or review posts for Medium, 英文博客, 微信公众号, 公众号, WeChat, weixin, 小红书, xhs, or Xiaohongshu while preserving provided facts and avoiding invented claims.