Coordinates development tasks with ten named skills and advisory input/output contracts, plus a linter for contract and workflow checks, without runtime enforcement.
adventurewave-labs/apsl-core
- Latest recorded source activity
- SkillsMP catalog refreshed
- skills collected
- 11
- GitHub stars
- 0
- GitHub forks
- 0
Install with an AI assistant
Copy this prompt into the AI assistant you're using.
Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/adventurewave-labs/apsl-core.Loading the repository README…
Posts
See all →Skills in this repository
Showing 11 of 11 collected skills.
Fixture: contract block with deliberate errors.
Launch a subagent to handle a specialized task in isolation, keeping the main context clean. Use when delegating research, browsing, or deep code analysis. Agent types: explore, browser, codebase, general.
Generate or update documentation for code, APIs, or architecture. Use when asked to document, write a README, add comments, or produce an ADR. Supports --type api|readme|inline|arch and --out filepath.
Stage and commit changes with a well-formed Conventional Commits message. Analyzes the diff to auto-generate a message if none is provided. Supports --msg "custom message", --no-push, --type feat|fix|chore|docs|refactor.
Generate boilerplate structure for a component, module, service, or project type, matching the project's existing conventions. Use when starting something new. Supports --name, --type, --framework, --lang.
Request a pause for explicit human approval before proceeding. Use as a checkpoint before irreversible operations like deploy, delete, publish, or any action with production impact.
Generate an explicit step-by-step plan before taking action. Use before complex, multi-step, or risky tasks to enforce think-before-act discipline. After presenting the plan, request a halt and wait for approval.
Diagnose and repair a bug, error, or failing test. Use when something is broken and needs root-cause repair. Supports --error "message", --auto, --test (run tests after fix).
Improve code quality, structure, or readability without changing observable behavior. Use for cleanup, DRY improvements, naming, extraction, reorganization. Supports --goal "description", --scope function|file|module, --auto.
Structured code or PR review with actionable feedback organized by severity. Use when asked to review changes, a branch, a PR, or specific files. Supports --type pr|security|performance|style|all.
Deep analysis of code, architecture, data, or any target. Use when asked to examine, audit, inspect, review internals, or understand something thoroughly. Supports --type security|perf|types|deps|arch and --format md|json.