🎯Skills46
An individual skill from the camacho/ai-skills collection, a set of AI skills for Claude Code and Codex installable via npx skills, with each skill providing focused instructions for specific development tasks.
An individual skill from the camacho/ai-skills collection, a set of AI skills for Claude Code and Codex installable via npx skills, with each skill providing focused instructions for specific development tasks.
An individual skill from the camacho/ai-skills collection, a set of AI skills for Claude Code and Codex installable via npx skills, with each skill providing focused instructions for specific development tasks.
An individual skill from the camacho/ai-skills collection, a set of AI skills for Claude Code and Codex installable via npx skills, with each skill providing focused instructions for specific development tasks.
An individual skill from the camacho/ai-skills collection, a set of AI skills for Claude Code and Codex installable via npx skills, with each skill providing focused instructions for specific development tasks.
A post-task review and memory extraction skill from a collection of AI skills for Claude Code and Codex. Part of a toolkit that includes validation, context reconstruction, and project naming capabilities, all installable via npx skills.
Part of a collection of AI skills for Claude Code and Codex that includes validation, post-task reflection, context reconstruction after /clear, and project naming. All skills are installed via npx skills with SKILL.md frontmatter discovery.
Creates high-quality skills by running three competing parallel approaches (skill-creator, superpowers, and manual), comparing results on five dimensions, then synthesizing the best elements into a final SKILL.md through a full 10-step workflow pipeline.
Exits copilot mode and returns Claude Code to autonomous mode with full worktree enforcement and the complete workflow pipeline active.
Handles session wrap-up by invoking reflection to extract learnings, cleaning up worktree branches (removing merged ones, warning about unmerged), and prompting the user to type /clear.
Reconstructs conversation context after a /clear command or session resume, helping AI agents quickly regain awareness of what was being worked on without the user needing to re-explain the project state.
Centralizes reviewer selection and loop governance for code review workflows using a frozen policy algebra that handles panel assembly, retention, expansion, and convergence decisions based on file types and severity gating.
Guides creating SessionStart hooks for Claude Code that install dependencies so tests and linters work across CLI, web, and API sessions, with support for async mode and environment variable persistence.
Runs full project validation including typechecking, linting, formatting, testing, and optional build steps to ensure code quality before commits or deployments.
Switches Claude Code into copilot mode where the human drives and Claude assists, relaxing worktree enforcement and allowing main branch commits with a sliding 4-hour TTL.
Generates frozen, drift-checkable Starlark governance rule blocks from intent strings or plan files, with adaptive deep/shallow interview modes and a --verify command for canonicalized drift detection against candidates.
Part of a collection of AI skills for Claude Code and Codex, installable via npx skills. The collection includes skills for template management, dotfile syncing, skill promotion, project validation, post-task review, context reconstruction, and project naming.
A universal task dispatcher skill that routes any development task through a structured 0-9 step workflow — handling surface selection, planning, implementation, review, and shipping while supporting both autonomous pipeline and human-directed copilot modes.
Part of a collection of AI skills for Claude Code and Codex, installable via npx skills. The collection includes skills for template management, dotfile syncing, skill promotion, project validation, post-task review, context reconstruction, and project naming.
A skill that automates merging a feature branch onto main via two paths: opening a PR with auto-merge when CI is up, or performing a local shallow-clone merge when CI is down. It validates the build before merging and handles worktree cleanup.
Part of a collection of AI skills for Claude Code and Codex, installable via npx skills. The collection includes skills for template management, dotfile syncing, skill promotion, project validation, post-task review, context reconstruction, and project naming.
Detects new or modified skills in the local .agents/skills/ directory by comparing git hashes against the ai-skills repo, then performs code review, publishes changes, and manages rollback snapshots.
Moves issues through a state-machine-driven triage workflow on the project issue tracker, generating agent briefs for AFK implementation and maintaining an out-of-scope knowledge base of rejected features.
Breaks a plan, spec, or PRD into independently-grabbable issues on the project issue tracker using tracer-bullet vertical slices, where each issue cuts through all integration layers end-to-end.
A curated AI agent skills collection for Claude Code and Codex, combining original skills with adapted community skills from the ecosystem.
Turns the current conversation context and codebase understanding into a PRD and publishes it to the project issue tracker, synthesizing existing context without interviewing the user.
A mandatory pre-implementation skill that guides collaborative brainstorming through structured dialogue, exploring project context, asking clarifying questions one at a time, proposing approaches with trade-offs, and producing an approved design document before any code is written.
A curated collection of AI agent skills for Claude Code and Codex, serving as a single source of truth for skills across all projects with adapted community skills.
A thin alias for the /task skill that serves as a universal task dispatcher, allowing users to invoke /task or /workflow interchangeably.
Writes comprehensive implementation plans from specs or requirements, producing bite-sized tasks with detailed guidance on files to touch, testing strategies, and documentation, assuming the engineer has zero codebase context.
A skill from a collection of AI skills for Claude Code and Codex, installable via npx skills, featuring project validation, post-task review, context reconstruction, dotfiles syncing, and template management capabilities.
Brings the current branch up to date with its parent branch using git merge (never rebase), with optional stack cascade to propagate changes through dependent branches. Designed as the inbound counterpart to /ship and /isolate.
A team position bootstrap skill that loads a persistent role identity from a position profile file, loads associated state (playbook and decisions), and runs the position-specific session-start sequence for roles like TPM, software engineer, or architect.
Curated collection of AI agent skills for Claude Code and Codex, serving as a single source of truth with forked community skills adapted for consistent project conventions.
Curated AI agent skills collection for Claude Code and Codex that provides consistent, project-adapted community skills with fixed references and package manager configurations.
A curated collection of AI agent skills for Claude Code and Codex, featuring original skills and adapted community forks with fixed paths, package manager references, and skill cross-references.
A planning skill that generates a structured blueprint (plan file) for multi-step tasks by running a brainstorming round and writing a plan from a template, including threat model selection, scope ceilings, and implementation steps.
Runs bounded review-fix-re-review cycles on plans or implementations with configurable severity gating (P0-P3), automatic fixes for above-gate findings, and panel governance via assemble-panel for reviewer selection and convergence.
Parks mid-session work and prints resume instructions, with a heaviness check that routes to memory consolidation if pending feedback or project notes need reflection.
A skill that handles Step 3 of the /task workflow — running one brainstorming round then writing a structured plan file with branch, target, threat model, scope ceiling, implementation steps, and confidence scaffold, and committing it before review.
A curated collection of AI agent skills for Claude Code and Codex, serving as a centralized source of truth with both original and forked community skills adapted for consistent project workflows.
A curated collection of AI agent skills for Claude Code and Codex, featuring forked and adapted community skills with fixed paths, package manager references, and cross-references for consistent workflow integration.
Skill
Skill
Skill
Skill