claude-code-harness

executor が返した advisor-request.v1 に対して方針だけ返す非実行 advisor

sprint-contract と review artifact を基準に verdict を返す read-only reviewer

実装、preflight 自己点検、検証、commit 準備を 1 タスク単位で進める統合ワーカー

Browser automation through the repo agent-browser CLI. Explicit helper for navigation, forms, screenshots, scraping, and web-app checks. Prefer Browser Use or Playwright when available. Do NOT load for: sharing URLs, embedding links, or editing screenshot files.

Explicit helper for authentication and payment implementation with Clerk, Supabase Auth, or Stripe. Do NOT load for: general UI work, database design, or non-auth features.

Team execution mode — backward-compatible alias for harness-work with team orchestration. Composer/composer 2.5 maps to the cursor backend.

Validates brainstormed ideas with Cursor PM, updates Plans.md, then handoff back. Cursor ↔ Claude Code 2-Agent workflow support. Use when user mentions Cursor PM handoff, 2-agent plan validation, CC-Cursor round trip, or brainstorm review. Do NOT load for: implementation work, single-agent tasks, or direct coding.

Quality guardrail for Claude/Codex update integration. Detects doc-only Feature Table additions and requires implementation or explicit planning. Internal use only.

CI red? Call us. Pipeline fire brigade deploys. Use when user mentions CI failures, build errors, test failures, or pipeline issues. Do NOT load for: local builds, standard implementation work, reviews, or setup.

Explicit helper for CRUD scaffolding and API endpoint generation. Do NOT load for: UI components, form design, database schema discussion, or general implementation.

Read-only delegate to cursor-agent (Composer) for questions, investigation, design discussion, and adversarial sanity checks. No worktree, no cherry-pick, no Lead diff review — cursor-agent is locked to ask mode and cannot write. Use when user says: ask cursor, cursor sanity check, get a second opinion, adversarial review, design discussion, investigate with cursor, cursor:ask. Do NOT load for: implementation, refactor, file edits, commit/push work, anything requiring write access (use cursor:do or breezing --cursor instead).

Delegate a single write task to Cursor Composer via cursor-companion.sh inside an isolated worktree, then Lead-review the diff and cherry-pick. Use when user invokes cursor:do, says delegate to cursor, have composer write it, refactor with cursor, hand a file edit to Composer. Do NOT load for: planning, code review only, read-only investigation, or multi-task team runs (use breezing --cursor or cursor:ask instead).

Diagnose and recover Cursor backend failures for Harness workflows. Use when user invokes cursor:rescue, Cursor delegation fails, cursor-agent is missing, setup-cursor fails, or backend resolution unexpectedly falls back to claude.

Run a Cursor Composer review as an advisory second opinion while keeping the primary review verdict on the host brain. Use when user invokes cursor:review, asks Cursor to review, or wants composer to sanity-check a diff. Cursor never owns APPROVE/REQUEST_CHANGES.

Configure and verify the Cursor backend for Claude Code Harness. Use when user invokes cursor:setup, wants Cursor as the local default implementation backend, or asks to check Cursor plugin/agent readiness. Distribution default remains opt-in; only local env/user settings are changed when explicitly requested.

Deploy to Vercel/Netlify. One-way ticket to production arranged. Use when user mentions deployment, Vercel, Netlify, analytics, or health checks. Do NOT load for: implementation work, local development, reviews, or setup.

Generate project intro slides with Nano Banana Pro. Internal/manual workflow only; use from an explicit /generate-slide command or a parent media workflow.

Auto-generate product demo videos. Internal/manual workflow only; use from an explicit /generate-video command or a parent media workflow. Requires Remotion setup.

Use gogcli for Google Workspace CLI operations (Drive/Sheets/Docs/Slides). Trigger when a user asks to check, list, search, export, read, or update Google files via gogcli; when a Google URL/ID needs parsing; when auth/account selection or safe read-only workflows are needed; or when troubleshooting gogcli access/errors. Do NOT load for: general file operations, non-Google cloud storage, or standard shell commands.

Generate an Acceptance Demo HTML for non-engineer vibecoders right before ship/wait/reject decision. Reads back the acceptance_criteria that were stored as personal-preference.v1 by harness-plan-brief (joined by user_request_hash), then renders a single-file HTML showing each criterion as verified or unverified along with a ship/wait/reject recommendation. Use when the user asks for an acceptance review, wants to decide whether to ship a delivered task, or says: acceptance demo, accept demo, 受け入れ判断, 受入レビュー, ship/wait/reject 判定, 検収レビュー. Do NOT load for: implementation, code review, release work.

Long-running task loop using /loop (Claude Code dynamic mode) and ScheduleWakeup to re-enter with fresh context on each wake-up. Internally invokes harness-work through Agent. Trigger: long-running, loop, wake-up, autonomous. Do NOT load for: one-shot task execution, review, release, planning.

Show how much this session/project orchestrated across backends (Claude / Codex / Cursor). Renders an on-demand HTML scorecard + terminal summary from the orchestration ledger and lifetime totals. Use when the user asks to see orchestration usage, a backend scorecard, which backend was used, how much Codex/Cursor was used, lifetime totals, or wants something to show off (自慢). Do NOT load for: implementing tasks, reviews, planning, or release.

Generate a Plan Brief HTML for non-engineer vibecoders before implementation starts. Searches harness-mem (project-only) for relevant past decisions, patterns, and Plans archive entries, then renders a single-file HTML artifact summarizing understanding, options, risks, acceptance criteria, and confidence. Use when the user requests a planning preview, a non-engineer-friendly summary before approval, or says: plan brief, planning preview, 計画概要, 計画レビュー. Do NOT load for: actual implementation, code review, release work.

HAR: Research-backed, team-validated task planning, Plans.md management, progress sync. Trigger: create a plan, add tasks, update Plans.md, mark complete, check progress. Do NOT load for: implementation, review, release.