Native-first workflow plugin for your AI Code and Codex: 16 task skills for planning, test-first implementation, debugging, verification, safe effects, cross-provider review, and orchestration, plus an optional concise output style, a destructive-command tripwire, and a self-diagnosis doctor.
Skills 16 in the Claude listing
Autonomous run Use when an approved goal must continue unattended across many steps, milestones, context resets, or sessions, including keeping a pull request moving until it is merge-ready. Design and plan Use to specify non-trivial behavior or interfaces, resolve requirements or tradeoffs, or produce a multi-step implementation plan. Evidence research Use to research a decision that needs evidence beyond the repository, such as current upstream behavior, historical rationale, cross-system incidents, or contested claims. Grill me Use when the user asks to be grilled, or wants a plan, decision, idea, or draft stress-tested through a structured interview before any action. Humanizing prose Use when the task itself is to draft, rewrite, edit, or preserve human-facing prose for clarity, attribution, or proportion; do not select merely because other work ends with a response. Independent review Use when security, auth, billing, concurrency, data integrity, or another high-stakes artifact needs adversarial review by a provider other than its author. Mcp setup Use when installing, configuring, debugging, or verifying an MCP server for a harness, or when a configured server's tools are missing or failing. Megapowers doctor Use when Megapowers seems inactive or broken: a skill did not fire, the output style is missing, a hook errors or times out, or the user asks to check the Megapowers setup. Memory hygiene Use when auditing, pruning, correcting, refreshing, or automatically applying approved evidence-based cleanup to native harness memory. orchestrating Use when two or more independent lanes can run in parallel, an output-only lane protects lead context from raw payload, or a handoff or durable run needs coordination. Safe effects Use when preparing a deploy, message, charge, migration, destructive query, DNS change, external API write, or any action with a real-world side effect. Systematic debugging Use when a bug, failing or flaky test, production incident, unexpected output, or performance regression has an unknown cause. Test first implementation Use when adding or changing behavior, fixing a confirmed bug, refactoring executable code, or implementing a feature with tests. Upgrading megapowers Use when asked to update, upgrade, refresh, reinstall, or migrate Megapowers, check for a newer release, or repair stale installed plugin state. Verify and finish Use to verify a completed outcome before claiming success, handing off, committing, merging, publishing, deploying, or closing a task. Writing agent instructions Use when creating, revising, or auditing an agent skill, AGENTS.md, CLAUDE.md, or nested project instructions, including trigger design, progressive disclosure, harness-aware scope, behavioral validat