Fast parallel codebase scouting — searches many directories at once, returning a file map + architecture fingerprint. Use before planning or cross-dir work. NOT for deep single-file reads.
中文简介正在准备中。当前显示原文说明。
Triages reviewer comments on a GitHub PR: verifies each against the codebase, then accepts/pushes back/clarifies, replies in-thread. NOT for reviewing (mk:review-pr); NOT for fixes (mk:fix).
中文简介正在准备中。当前显示原文说明。
Reviews a GitHub PR with a shallow correctness/security/breaking/AI-slop checklist and emits a verdict; optionally posts via gh. NOT for own-diff audit (mk:review); NOT for replying (mk:respond-pr).
中文简介正在准备中。当前显示原文说明。
Domain-aware complexity routing: scans the task for domain keywords, matches domain-complexity.csv, and returns complexity level, workflow intensity, and execution tier. Extends Phase 0 orchestration.
中文简介正在准备中。当前显示原文说明。
Post-build pass reducing complexity without changing behavior. Catches dead code, needless abstractions, over-engineering. NOT for style (mk:clean-code) or diff review (mk:review).
中文简介正在准备中。当前显示原文说明。
Pre-ticket Fibonacci sizing for pasted stories; advisory by default, writes a Sizing Report. Opt-in auto-create adds a confirm gate. NOT for single-ticket sizing (mk:jira-estimator).
中文简介正在准备中。当前显示原文说明。
Negotiates a file-based sprint contract between generator and evaluator before code is written: turns a spec into testable acceptance criteria with rubric tie-ins.
中文简介正在准备中。当前显示原文说明。
Generate visual artifacts: explanations, diagrams, slides, diff views. NOT plan-as-HTML (mk:visual-plan), live media gen (mk:multimodal), or browser QA (mk:qa).
中文简介正在准备中。当前显示原文说明。
Correct file paths, project layout, and naming conventions. Two modes: advisory (return a path) and organize (restructure).
中文简介正在准备中。当前显示原文说明。
Coordinate staged or parallel jobs across coding-agent runtimes and subagents: routes by capability/risk, isolates writers in worktrees, arbitrates results. NOT for a single-agent task (mk:cook).
中文简介正在准备中。当前显示原文说明。
Structured multi-file implementation plans before build; scope-aware, enforces Gate 1. NOT for ticket sizing (mk:planning-engine) or plan CEO-review (mk:plan-ceo-review).
中文简介正在准备中。当前显示原文说明。
Use when stuck on approach, not debugging. Unstick via 7 techniques: simplification, collision-zone thinking, inversion, first principles, via negativa. For debugging use mk:sequential-thinking.
中文简介正在准备中。当前显示原文说明。
Analyzes ticket complexity and dependencies against the codebase before sprint planning. NOT for writing plans (mk:plan-creator); NOT for scope review (mk:plan-ceo-review).
中文简介正在准备中。当前显示原文说明。
Session-persistent browser automation via Playwright CLI: form filling, screenshots, data extraction, multi-step flows. NOT for autonomous flows (mk:agent-browser) or .spec.ts E2E code (mk:qa-manual).
中文简介正在准备中。当前显示原文说明。
Refine a draft prompt before sending to a coding agent: decomposes goal/context/constraints, flags ambiguity, emits a model-agnostic rewrite. NOT for prompts from scratch (mk:brainstorming).
中文简介正在准备中。当前显示原文说明。
Systematically QA-tests a web app and fixes bugs: test, then iteratively fix, re-verify, commit each fix. Tiers: Quick, Standard, Exhaustive. Produces before/after health scores and fix evidence.
中文简介正在准备中。当前显示原文说明。
Resolves an in-progress git merge/rebase conflict: reads conflict state, recovers each side's intent, resolves hunks, runs checks. NOT for the ship pipeline (mk:ship); NOT for bug debugging (mk:fix).
中文简介正在准备中。当前显示原文说明。
Weekly engineering retrospective from commit history, work patterns, and code-quality metrics, with trend tracking and per-person breakdown. NOT for per-session reflection (mk:memory / Phase 6).
中文简介正在准备中。当前显示原文说明。
Loads, lists, composes, or validates graded evaluation rubrics: PASS/WARN/FAIL grading with weighted criteria and balanced anchor examples for evaluator-style review.
中文简介正在准备中。当前显示原文说明。
Ship workflow: prepare stages/commits locally; release pushes + opens a PR; publish handles issues + versioning. Use for 'ship', 'push', 'create a PR'. Official (→main) and beta (→dev) modes.
中文简介正在准备中。当前显示原文说明。
Persists and restores mid-session workflow progress across context resets. Use when resuming workflows, saving state, or handling context compaction. Triggers on: handoff, save state, resume workflow.
中文简介正在准备中。当前显示原文说明。
Task claiming and file-ownership tracking for parallel execution: agents claim tasks from a shared queue with ownership enforced. Use during parallel phases with multiple simultaneous agents.
中文简介正在准备中。当前显示原文说明。
Analyzes harness trace logs to surface recurring failure patterns and suggest harness improvements. Scatter-gather across last N runs with a mandatory HITL gate before any change is applied.
中文简介正在准备中。当前显示原文说明。
Pack an EXTERNAL repository into one AI-friendly file (markdown/xml/json) for third-party analysis, security audits, or handoff. Do NOT use for the current project — it's already read lazily.
中文简介正在准备中。当前显示原文说明。