Use when creating, consolidating, editing, testing, or reviewing coding-agent skills, skill packs, orchestration skills, reusable software-engineering workflows, or instructions meant to steer coding-agent behavior in repos.
EnglishUse when the user asks for cleanup beyond incidental touched-file polish: explicit human-reviewability cleanup for comments/docstrings/imports/code order, cleanup discovery, continued cleanup loops, behavior-preserving refactors with structure risk, duplicated logic, module-boundary or caller-interface simplification, testability blockers, or obsolete-code removal.
EnglishUse when starting nontrivial repo work where route choice, scope, assumptions, or completion evidence could drift; defaults to a micro-loop and prevents skill pileups.
EnglishReview local PRs and high-risk local diffs through independent fresh-context reviewers and a root-verified evidence ledger. Use for release gates, shared plumbing, migrations, security or permission changes, CI/workflow/config, public interfaces, and data contracts. Hand off ordinary fixed-point Standards/Spec review to $review.
EnglishUse when repo work needs durable GitHub recordkeeping: issue/PRD/slice bodies, claim/status/blocker comments, PR/issue links, readiness fields, or CI/review evidence recorded in GitHub; skip ordinary local work, generic triage, PR publishing, CI debugging, and review-comment resolution when narrower workflows own them.
EnglishUse when approved repo work should be coordinated through a plan document, GitHub issues, one verified implementation or research issue at a time, and an explicit checkpoint policy.
EnglishDrain one parent spec or PRD's ready implementation ticket graph through serial or parallel isolated lanes, serial integration, root-owned review, Lock, and verified child and parent closeout.
EnglishUse when approved multi-step repo work truly needs a written execution plan or handoff with task boundaries, checks, execution mode, ownership scope, continuity, and final verification; skip tiny edits, one behavior, one bug, or issue-only/PR-only tracking.
EnglishUse when the user explicitly asks for subagents, parallel agents, worker agents, independent code review, multi-agent codebase exploration, subagent-ready handoffs, or an approved plan/issue calls for authorized, bounded delegation that is cheaper than local work.
EnglishUse when approved work changes caller-visible or user-visible behavior, adds a new behavior path, or an understood bug needs a new durable behavior or regression check.
EnglishUse as a lightweight gate when overlapping dirty tracked paths, relevant untracked protected paths, or repo operations could overwrite, delete, publish, mix, stage, regenerate, or depend on unknown work: staging, commit, PR, push, merge, branch/worktree changes, installs, lockfiles, migrations, generators, broad formatters, generated cleanup, or destructive Git/filesystem commands.
English
AGENT SKILLS ATLASIncludes Author Skills, Codebase Cleanup, Coding Router, and other workflows.
Claude Code
Codex
Cursor
GitHub Copilot
Gemini CLISupported by the Atlas install prompt. Skill-specific requirements vary.
Pack stats
What’s included
Author Skills
Use when creating, consolidating, editing, testing, or reviewing coding-agent skills, skill packs, orchestration skills, reusable software-engineering workflows, or instructions meant to steer coding-agent behavior in repos.
Codebase Cleanup
Use when the user asks for cleanup beyond incidental touched-file polish: explicit human-reviewability cleanup for comments/docstrings/imports/code order, cleanup discovery, continued cleanup loops, behavior-preserving refactors with structure risk, duplicated logic, module-boundary or caller-interface simplification, testability blockers, or obsolete-code removal.
Coding Router
Use when starting nontrivial repo work where route choice, scope, assumptions, or completion evidence could drift; defaults to a micro-loop and prevents skill pileups.
Convergent PR Review
Review local PRs and high-risk local diffs through independent fresh-context reviewers and a root-verified evidence ledger. Use for release gates, shared plumbing, migrations, security or permission changes, CI/workflow/config, public interfaces, and data contracts. Hand off ordinary fixed-point Standards/Spec review to $review.
GitHub Tracking
Use when repo work needs durable GitHub recordkeeping: issue/PRD/slice bodies, claim/status/blocker comments, PR/issue links, readiness fields, or CI/review evidence recorded in GitHub; skip ordinary local work, generic triage, PR publishing, CI debugging, and review-comment resolution when narrower workflows own them.
Issue Driven Execution
Use when approved repo work should be coordinated through a plan document, GitHub issues, one verified implementation or research issue at a time, and an explicit checkpoint policy.
Parallel Implement
Drain one parent spec or PRD's ready implementation ticket graph through serial or parallel isolated lanes, serial integration, root-owned review, Lock, and verified child and parent closeout.
Slice Plan
Use when approved multi-step repo work truly needs a written execution plan or handoff with task boundaries, checks, execution mode, ownership scope, continuity, and final verification; skip tiny edits, one behavior, one bug, or issue-only/PR-only tracking.
Subagent Workflow
Use when the user explicitly asks for subagents, parallel agents, worker agents, independent code review, multi-agent codebase exploration, subagent-ready handoffs, or an approved plan/issue calls for authorized, bounded delegation that is cheaper than local work.
TDD Slice
Use when approved work changes caller-visible or user-visible behavior, adds a new behavior path, or an understood bug needs a new durable behavior or regression check.
Workspace Safety
Use as a lightweight gate when overlapping dirty tracked paths, relevant untracked protected paths, or repo operations could overwrite, delete, publish, mix, stage, regenerate, or depend on unknown work: staging, commit, PR, push, merge, branch/worktree changes, installs, lockfiles, migrations, generators, broad formatters, generated cleanup, or destructive Git/filesystem commands.
Quick start
- Choose your skills. Select individual skills above, or choose the entire pack.
- Copy the install prompt. Use Copy install prompt above the list.
- Send it to your agent. Paste into your coding agent’s chat. Your agent reviews the source and guides the setup.
Try it with your agent
After installation, start with a request like:
Use the Author Skills skill to help with my task.
Ask me for the context you need before getting started.Requirements & license
- A coding agent that can access your project files and install skills.
- Your agent checks each skill’s instructions, dependencies, and permissions during setup.
- Reported license for the skills shown: MIT. The author’s terms apply.