Comments
Code comments and annotations
docs-codebase
by vasilyu1983
Technical writing patterns for README files, API documentation, architecture decision records (ADRs), changelogs, contributing guides, code comments, and docs-as-code workflows. Covers documentation structure, style guides, Markdown best practices, and documentation testing.
project-astrology-tarot-divination
by vasilyu1983
Use when interpreting Tarot cards, designing spreads, performing readings, or building AI tarot features (daily card, card combinations, timing). Covers Major/Minor Arcana meanings, spread layouts (incl. Celtic Cross), reading ethics, and safe AI positioning. Not for medical/legal/crisis support or death/illness predictions. (project)
evaluate-enrichment
by monarch-initiative
Use when evaluating a GO enrichment method or a GOA evidence variant (all vs IBA vs IBA+IEA vs no-contributes_to) against the curated benchmark in this repo — the IBA/evidence-ablation eval. Covers building queries.gmt, prepare_go_eval, running genesets-rs, and scoring with the confirmatory/mechanistic split and the recovery_status diagnostics, plus the don't-refit-the-gold guardrail.
triage
by lirtual
Move issues and external PRs through a state machine of triage roles, categorise, verify, grill if needed, and write agent-ready briefs.
curate-geneset
by monarch-initiative
Use when curating or reviewing a GO-term interpretation of a non-GO gene set (MSigDB C8/C2/H, literature disease-activity, or GWAS/CRISPR genetics) for the curation/ gold standard in this repo. Covers grounding in the real membership, OLS term verification, the category / recovery_status / insight axes, evidence, series, the 4-gate validator, and folding the set into the eval.
triage
by agbossjs
Conveyor stage 1 → status Triage. Takes a requirement (or an existing issue), clarifies its business side through quizzes — why, what it does, what it does NOT do, constraints, business acceptance criteria (AC1…) — and creates or updates GitHub issues, splitting into a parent with sub-issues when useful. Use whenever the user brings a new requirement, idea, request, bug report or an issue number/URL and wants it clarified, triaged, "заведи тикет", "разбери задачу" — before any technical spec or code.
spec-session
by agbossjs
Conveyor stage 2 → status Backlog. For issues in Triage, decides the technical side through quizzes and writes into the issue a detailed spec, an implementation plan, acceptance criteria split into automated (tests) and manual (plain-language instructions for a human — where to go, what to do, what to see), test cases, wiki pages to update and dependencies. Use after triage, or whenever the user says "spec", "спека", "план имплементации", "распиши тикет", or wants issues ready for autonomous implementation.
pair-mode
by ElliotJLT
When activated, Claude becomes a pair programming partner rather than a code-generation tool. Explains reasoning, teaches patterns, asks questions that build understanding, and adjusts depth based on the person's goals. Balances getting work done with building lasting knowledge. Activate with "let's pair on this" or by setting a learning focus at session start.
docx
by adamwstauffer
"Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files). Triggers include: any mention of \"Word doc\", \"word document\", \".docx\", or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a \"report\", \"memo\", \"letter\", \"template\", or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation."
youtube
by agentproto
Drive the user's authenticated YouTube via the browser skill: channels, videos, comments + commenters, search via the InnerTube API; gated like/comment/subscribe. Use for YouTube research/automation as the user.
tiktok
by agentproto
Drive the user's authenticated TikTok via the browser skill: feed/user/comments via the /api endpoints (signing-gated — read via page responses). Use for TikTok research as the user.
by agentproto
Drive the user's authenticated Instagram via the browser skill: profiles, media likers/comments, followers/following, search via the web private API; gated like/comment/follow/DM. Use for Instagram research/outreach as the user.
sql-optimization
by chaterm
SQL 优化与调优
code-review
by letta-ai
Skill for handling PR code reviews. Use when triggered by a PR review comment, review request, or when asked to review code changes. Provides workflow for reading review comments, understanding feedback, and iterating on changes.
security-fencing
by KingInYellows
"Canonical prompt-injection hardening block for agents that analyze untrusted content (source code, CI logs, workflow files). Use when authoring an agent that reads untrusted input — copy the appropriate variant verbatim into the agent body; only pure output-only agents (no file reading, no untrusted input) are exempt."
plan-status
by KingInYellows
Show a per-file checkbox progress dashboard of plans/ (open) and plans/complete/ (archived) plan documents. Use when reviewing which plans are ready to archive or checking work-in-flight at a glance.
local-config
by KingInYellows
"Defines the yellow-plugins.local.md per-project config file schema. Use when authoring commands that should accept per-project overrides for review pipeline behavior, reviewer set narrowing, depth controls, or focus-area filtering."
co-review
by taloncjones
Freeze a finished change, collect two or four independent review seats by change class, and evaluate one fail-closed final-gate report.
plan-status
by KingInYellows
'Show a per-file checkbox progress dashboard of plans/ (open) and plans/complete/ (archived) plan documents. Use when reviewing which plans are ready to archive or checking work-in-flight at a glance.'
implement
by agbossjs
Conveyor stage 3 → status In progress, and Test after merge. Implements Backlog issues — in parallel where they are independent (git worktrees with sub-agents, or separate Claude Code on the web sessions), in dependency order otherwise — each as its own branch and PR into main with tests, green gate, self-review, Wiki update and a manual-test handoff. Never moves anything to Done. Use after spec-session, or when the user says "implement", "имплементируй", "запусти", "go", or names Backlog issues to build.
debugging
by KingInYellows
"Find root causes systematically before fixing — trace the causal chain and form hypotheses with predictions for uncertain links. Use when debugging errors, test failures, stack traces, or when stuck after failed fix attempts. For delegating a stuck investigation to an independent Codex session, use /codex:rescue."
stack-decomposition-format
by KingInYellows
Structured format contract for "## Stack Decomposition" sections in plan documents, shared by gt-stack-plan (producer) and flow:work (consumer). Use when writing or parsing a Stack Decomposition section.
ce-resolve-pr-feedback
by KingInYellows
Resolve PR review feedback by evaluating validity and fixing issues in parallel. Use when addressing PR review comments, resolving review threads, or fixing code review feedback.
dotfiles-change-control
by taloncjones
Load BEFORE making, committing, or reviewing ANY change in the dotfiles repo. Covers change classification (tracked asset vs installer logic vs machine-local template vs vendored-untracked vs project .claude config), the full gate chain (guard hooks, bin/dotfiles-tests, public-safety scan, CI), commit/branch/PR rules, and the non-negotiables with their incident history. Trigger phrases -- "commit this", "open a PR", "why did my commit get blocked", "can I edit settings.json.tmpl", "can I edit these rules files", "hook blocked me", "bypass the hook". NOT for diagnosing why something is broken (dotfiles-debugging-playbook), running the tests as evidence (dotfiles-validation-and-qa), or historical why-decisions (dotfiles-failure-archaeology).