メニュー
Initialize a new project with Harness engineering discipline. Creates directory structure, copies templates, sets up Makefile, installs skills. Use when starting a new project or adopting Harness on an existing codebase.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| name | harness-init |
| description | Initialize a new project with Harness engineering discipline. Creates directory structure, copies templates, sets up Makefile, installs skills. Use when starting a new project or adopting Harness on an existing codebase. |
Bootstrap a project with the full Harness engineering stack: spec governance, DDD enforcement, TDD role isolation, and blast-radius-based autonomy.
{project_root}/
├── .harness/
│ └── config.yaml
├── .pm/
│ ├── stable/
│ ├── runtime/
│ └── design/
│ └── generated-concepts/
├── specs/
├── maintenance/
│ ├── debug/
│ └── index.md
├── docs/
│ └── adr/
├── CONTEXT.md
├── CLAUDE.md
├── AGENTS.md
└── Makefile
| Template | Destination | Purpose |
|---|---|---|
Makefile | ./Makefile | Gatekeeper targets (verify, verify-ai) |
CONSTITUTION_TEMPLATE.md | ./CONSTITUTION.md | Project principles |
AGENTS.md | ./AGENTS.md | Cross-agent standards |
CLAUDE.md | ./CLAUDE.md | Claude-specific routing layer |
CACHE.md | ./CACHE.md | Cache geometry protocol (stable-first context order) |
| Source | Destination | Purpose |
|---|---|---|
scripts/verify-ai.sh | ./scripts/verify-ai.sh | TDD role boundary enforcement |
scripts/classify-risk.sh | ./scripts/classify-risk.sh | Path-based blast radius classifier |
CONTEXT.md — empty glossary stub with the format from ../domain-language/CONTEXT_FORMAT.mdUBIQUITOUS_LANGUAGE.md — empty DDD term mapping stub.harness/config.yaml — project config (project name, risk policy path, skill paths)specs/ — empty directory for feature specsmaintenance/debug/ — empty directory for debug recordsmaintenance/index.md — empty maintenance index (from DEBUG_INDEX_TEMPLATE.md)docs/adr/ — empty directory for architecture decision records--pm flag)When --pm is specified, additionally create:
.pm/stable/ — copy all templates from references/templates/pm/ (product.md, evidence.md, value-proposition.md, ux-principles.md, user-journeys.md, ui-direction.md, roadmap.md, stage-definitions.md, architecture-guardrails.md, acceptance-rubric.md).pm/runtime/ — copy runtime templates (state.yaml, active-stage.md, next-task.md, worker-report.md, acceptance-review.md, spike-report.md, blockers.md, loop-log.md, handoff.md, loop-control, worker-config.yaml).pm/design/ — empty directory for design probes and UI feedback.pm/decisions.md — decision log template.pm/runtime/loop-control — initialized with CONTINUECheck if this is a greenfield project (empty directory) or an existing codebase. If existing, scan for conflicts with existing files before overwriting.
Templates live in the Harness repo at templates/. If running from an installed skill, resolve relative to the skill's install location. For details, see INIT_GUIDE.md.
Create the directory tree. Copy templates. Copy Harness scripts (verify-ai.sh, classify-risk.sh) to ./scripts/. Generate skeleton files.
Never overwrite existing files. If a file already exists (e.g., the project already has a Makefile), print a warning and skip. The user can merge manually.
Run harness install-skills (or equivalent) to wire skill references into the agent's config.
✅ Harness initialized. Next steps:
1. Fill CONTEXT.md with your domain terms (use harness-domain-language)
2. Write your first spec: specs/001-{feature}/spec.md
3. Debug issues go to: maintenance/debug/
4. Run make verify to check setup
--force — overwrite existing files (use with caution)--minimal — only create CONTEXT.md, docs/adr/, and Makefile (skip templates)--pm — additionally create .pm/ directory with all PM file templates for product evolution--templates-dir <path> — use custom template directory instead of built-inFor the full initialization guide including template resolution and post-init workflows, see INIT_GUIDE.md.
If a project already has Harness but needs PM support, harness-grill-product will create missing .pm/ files on first run. Alternatively, manually create the structure:
mkdir -p .pm/stable .pm/runtime .pm/design/generated-concepts
# Copy templates from references/templates/pm/ into .pm/stable/ and .pm/runtime/
Entry point and router for Harness-governed AI engineering. Use automatically for software engineering tasks, feature implementation, bug fixes, refactors, architecture changes, repository initialization, spec/plan/task workflows, risk classification, TDD, evaluation, reporting, debugging, product definition, supervisor PM loops, code semantic auditing, or when the user says "use Harness", "按 Harness 流程", "接管这个任务", "implement this", "fix this", "debug this", "review this", "product idea", "worth building", "goal", "supervisor", "atlas", "语义地图", "语义审计", "audit this code".
Low token Codex PM loop skill. Reads .pm/ files, generates bounded tasks, delegates to OpenCode Intern, reviews reports, and updates project state. Designed for long-running /goal sessions where Codex acts as product supervisor and OpenCode acts as engineering intern. Use when: "goal", "supervisor", "PM loop", "advance the product", "long-running product evolution", "Codex /goal".
Convert source code into semantic audit documentation — Mermaid diagrams, structured tables, and pseudocode — for human semantic verification. Triggers on "atlas", "semantic map", "audit this code", "代码语义地图", "语义审计".
OpenCode CLI 调用指南。提供所有 CLI 命令、标志、环境变量的速查参考，用于在终端中编程式调用 OpenCode。当用户需要通过 shell 脚本、CI/CD、自动化流水线或非交互模式使用 OpenCode 时触发。触发词：opencode cli、opencode run、opencode serve、非交互模式、自动化调用、脚本调用 opencode。
OpenCode worker skill. Executes bounded tasks written by harness-supervisor. Reads next-task.md, follows Harness engineering flow, modifies code, writes tests, runs verification, and outputs structured worker-report.md. Use when: "execute task", "run intern", "worker", or when instructed by supervisor to read .pm/runtime/next-task.md.
High-level product definition and UX discovery skill. Runs seven gates to determine whether a product is worth building, who it's for, what experience should exist, what the MVP boundary is, and what UI/UX direction should guide implementation. Produces .pm/stable/ files and initializes .pm/runtime/state.yaml. Use when: product idea, "worth building", "product definition", "define this product", user journey, UX principles, UI direction, MVP boundary.