Agent Skills by m2ai-jp
Showing 18 vetted skills indexed across 1 GitHub repositories.
state
Manage state.md updates and orchestrate playbook operations.
golden-path
Generate a playbook skeleton via the Hook, Skill, and pm SubAgent chain.
playbook-init
Delegate new task requests to the pm SubAgent for playbook creation and state updates.
playbook-gate
Block Edit/Write actions when playbook.active is null across executors.
lint-checker
Run ESLint and TypeScript checks on JavaScript files and apply fixes.
frontend-design
Create production-ready frontend interfaces with typography, color, and motion guidelines.
context-management
Manage session history with context trimming and structured summaries.
abort-playbook
Abort a playbook workflow and archive it with aborted status.
access-control
Blocks edits to protected files and dangerous Bash commands based on state-driven rules.
understanding-check
Clarify ambiguous task intent using a 5W1H framework before execution.
deploy-checker
Automate deployment readiness checks for CI/CD and Vercel projects.
git-workflow
Automate GitHub PR creation, merging, and main branch synchronization.
quality-assurance
Automate linting, integrity checks, and health monitoring in Git workflows.
session-manager
Manage Claude session lifecycles with start, end, and snapshot workflows.
post-loop
Release Edit/Write blocks and derive the next playbook after completion.
plan-management
Coordinate hierarchical planning and playbook management for long-running agent sessions.
test-runner
Automate test execution and result reporting for software projects.
reward-guard
Validates state transitions with technical, consistency, and completeness checks before marking tasks done.