impl-project-structure
Generate the PROJECT_STRUCTURE manifest — annotated source tree, module boundaries, import conventions, and ADR-decided stack assumptions (language/framework-agnostic). Optional tool a developer may use anytime when establishing or extending the project skeleton; not required before code (TASK-230, 2026-09-16 human decision). Vendored Spine skill; see src/manifest-policy.js.
- License 许可
- MIT
- 最近更新
- 2026-10-10
- Stars
- ★ 1
- Forks
- 0
01 前言 Front Matter
- name
- impl-project-structure
- description
- Generate the PROJECT_STRUCTURE manifest — annotated source tree, module boundaries, import conventions, and ADR-decided stack assumptions (language/framework-agnostic). Optional tool a developer may use anytime when establishing or extending the project skeleton; not required before code (TASK-230, 2026-09-16 human decision). Vendored Spine skill; see src/manifest-policy.js.
03 SKILL.md 正文
Sources & manifest status (hivemind)
Vendored from implementation-engine. In hivemind the "context" this skill reads is the project's
knowledge surface — PROJECT.md, the wisearcher brain graph (kb_search/kb_answer, or the local
knowledge/ KB when the brain is offline), and the ticket's scope + acceptance criteria. Where the
body below refers to context/<file>.md, read the equivalent from that surface. Output goes to
implementation/PROJECT_STRUCTURE.md.
Status (TASK-230, 2026-09-16 human decision): optional tool, not a gate. No verification_tier
requires this manifest before code — see src/manifest-policy.js (requiresManifest always returns
false). Generate/update it whenever it helps, in whatever order the Developer chooses; the
reviewer does NOT treat a missing manifest as a finding. Preserve epistemic markers
([EXPLICIT]/[INFERRED:strong|weak]/[ASSUMED]) and respect source tiers.
impl-project-structure
Generates implementation/PROJECT_STRUCTURE.md — an annotated source tree with module boundaries and import conventions.
Read first
Read these files from context/ (and only from context/):
context/technical.md— tech stack, system components, architecture patternscontext/scope.md— feature modules and their boundariescontext/adrs.md— all ADR decisions (especially framework, navigation, state, API)
Output: implementation/PROJECT_STRUCTURE.md
Section 1: ADR decisions summary
Open with a compact table of all ADR decisions from context/adrs.md. These are the project-specific technology choices that inform the structure below.
## Project Technology Decisions
| ADR | Decision | Status |
|-----|----------|--------|
| ADR-01 | Mobile framework: React Native | Confirmed |
| ADR-04 | State: React Query + Zustand | Confirmed |
| ... | ... | ... |
Section 2: Source tree
Present the full source tree with an inline comment per file/folder explaining its purpose. Base the structure on the decided stack (ADR-01 through ADR-07) but add a note at the top that the logical boundaries (screens/, services/, state/) apply regardless of language or framework.
src/
├── navigation/ # router setup; one file per navigator
│ ├── RootNavigator # entry point: auth gate → main tabs
│ ├── AuthNavigator # login, register, onboarding stack
│ ├── MainTabNavigator # 4-tab bottom nav
│ ├── CampaignNavigator # campaign list → detail → workflow stack
│ └── ProfileNavigator # profile, settings stack
│
├── screens/ # one folder per scope item (S-##)
│ ├── auth/ # S-AUTH: login, register, onboarding, forgot-password
│ ├── home/ # S-HOME: dashboard
│ ├── campaigns/ # S-CAMP: list, filter, detail, apply
│ ├── workflow/ # S-WORK: 7-step campaign workflow
│ ├── chat/ # S-CHAT: brand threads
│ ├── profile/ # S-PROF: metrics, mediakit, campaign history
│ ├── settings/ # S-SET: personal data, security, preferences
│ ├── payments/ # S-PAY: balance, history, collect
│ └── reports/ # S-REP: analytics, CSV export
│
├── components/ # shared components (see COMPONENT_CATALOG.md)
│ ├── primitives/
│ ├── forms/
│ ├── navigation/
│ ├── data-display/
│ ├── feedback/
│ └── layout/
│
├── services/ # external integrations (one file per integration)
│ ├── http/ # HTTP client setup, auth interceptors, error handling
│ ├── push/ # push notification registration and handling
│ ├── storage/ # secure token storage abstraction
│ └── analytics/ # crash reporting + usage analytics
│
├── queries/ # server state: query hooks grouped by module
│ ├── auth.ts
│ ├── campaigns.ts
│ ├── workflow.ts
│ ├── payments.ts
│ ├── profile.ts
│ ├── chat.ts
│ └── reports.ts
│
├── state/ # UI state stores (one file per domain)
│ ├── ui.ts # global UI flags (loading overlays, active tab)
│ └── <module>.ts # per-module UI state if needed
│
├── i18n/ # internationalization
│ ├── index # i18n initialization, language detection
│ ├── locales/
│ │ ├── es.json # Spanish (primary)
│ │ └── en.json # English (confirmed in scope)
│
├── types/ # shared TypeScript interfaces
│ ├── api.ts # request/response shapes
│ ├── domain.ts # Campaign, Profile, Payment, etc.
│ └── navigation.ts # route params per navigator
│
└── utils/ # pure helpers (no side effects)
├── formatters.ts # currency, dates, numbers
├── validators.ts # RUC, phone, email
└── constants.ts # app-wide constants
Section 3: Module boundaries
Define what each module owns and what it must not import from other modules:
screens/may import fromcomponents/,queries/,state/,services/,i18n/,types/,utils/queries/may import fromservices/http/andtypes/state/may import fromtypes/onlycomponents/may import fromtypes/,i18n/,utils/— never fromscreens/orqueries/services/must not import fromscreens/,queries/, orstate/
Section 4: Adaptation note
Add a brief note explaining that the logical boundaries (screens, services, queries/data-fetching, state, components, i18n, types, utils) are framework-agnostic. The file names and tool choices in Section 2 reflect the ADR decisions for this project; a different stack would use the same boundaries with different tooling.
Rules
- Base file/folder names on the decided stack only in sections that call out ADR decisions explicitly.
- Keep the tree focused on
src/— do not enumerate config files, CI scripts, or test infrastructure here. - Do not add files for out-of-scope features (AI agent integration, backend, web platform).
After writing
Report: number of top-level modules, number of screens folders, any structural decisions that needed [ASSUMED] or [MISSING_INFO] markers.
04 同仓库更多 Skills
unknown-skill
A demo skill with no license signal at all, used to exercise the "unknown" classification branch of the human license gate (TASK-135).
spoofed-license-skill
A demo skill used to test TASK-150's SPDX-header-vs-LICENSE-file conflict surfacing.
permissive-skill
A demo skill used to test assimilation of a permissively-licensed third party skill.
malicious-skill
A fixture skill deliberately carrying risky content patterns, used only to exercise src/skill-scan.js (TASK-122). Never assimilated for real.
copyleft-skill
A demo skill used to test the human license gate for a copyleft third party skill.
watch
Watch a video (URL or local path). Downloads with yt-dlp, extracts auto-scaled frames with ffmpeg, pulls the transcript from captions (or Whisper API fallback), and hands the result to Claude so it can answer questions about what's in the video. (assimilated for Hivemind)