Skip to content

docs(scaffold): fill the repo's own AGENTS.md / CLAUDE.md / .specs placeholders (closes #99)#111

Merged
wesleysimplicio merged 1 commit into
mainfrom
claude/issue-99-fill-placeholders
Jun 2, 2026
Merged

docs(scaffold): fill the repo's own AGENTS.md / CLAUDE.md / .specs placeholders (closes #99)#111
wesleysimplicio merged 1 commit into
mainfrom
claude/issue-99-fill-placeholders

Conversation

@wesleysimplicio
Copy link
Copy Markdown
Owner

Resumo

Fecha #99. Preenche os placeholders nos arquivos canônicos do próprio scaffold (que existe pra remover placeholders de hosts).

Mudanças

  • AGENTS.md, CLAUDE.md, .github/copilot-instructions.md — substituem <STACK> e lista de <APP_NAME>/<FRONTEND_URL>/<BACKEND_URL>/<DATABASE_REQUIREMENT>/<AUTH_FLOW>/<EVIDENCE_COMMAND> por:
    • Identidade real (CLI tool, sem UI, sem DB, sem auth, PyPI-only desde 0.7.x).
    • Stack real (Python 3.10+ canônico + Node mirror + Rust opt-in + Playwright).
    • Bloco de comandos com o que de fato roda no repo (python -m unittest discover, ruff check, node scripts/check-version-sync.js, (cd rust && maturin develop)).
  • .specs/product/VISION.md — tese real do produto.
  • .specs/product/DOMAIN.md — entidades reais (ProjectFile, Precedent, Module/Layer, Symbol, CallEdge, Endpoint, Screen, IndexState, Receipt) + 7 regras críticas + edge cases.
  • .specs/product/PERSONAS.md — 5 consumidores reais (SendSprint orchestrator, simplicio-dev-cli, humano de manutenção, CI/DoD, maintainer de release).

Zero mudança de código.

Acceptance

  • Nenhum <STACK> / <APP_NAME> / <PRODUCT_NAME> / <DOMAIN> em AGENTS.md, CLAUDE.md, .specs/product/ (verificado por grep).
  • Comandos listados rodam de fato (npm run lint ✅ 0 erros, python -m unittest discover ✅ 30 pass, ruff check ✅).
  • .github/copilot-instructions.md reflete a mesma stack/comandos.

Refs #99.

https://claude.ai/code/session_01JdmemqddwFnvbceWyuDE8m


Generated by Claude Code

…pecs/product

Closes #99.

This repo is the canonical scaffold whose whole purpose is to remove
placeholders from host projects, yet its own AGENTS.md and CLAUDE.md still
carried `<STACK>` lines and the host-only placeholder list (`<APP_NAME>`,
`<FRONTEND_URL>`, `<BACKEND_URL>`, `<DATABASE_REQUIREMENT>`, `<AUTH_FLOW>`,
`<EVIDENCE_COMMAND>`), and the commands block still pointed at generic
`npm run dev` / `npm run build` that do not exist here. The `.specs/product/`
files were also template prose.

Fill them with the repo's real identity:

- AGENTS.md, CLAUDE.md, and .github/copilot-instructions.md now declare the
  real stack (Python 3.10+ canonical with `orjson` + `diskcache`, Node CLI
  mirror at `bin/cli.js`, optional Rust/PyO3 acceleration, Playwright E2E),
  list the actual project-identity slots (CLI tool, no UI, PyPI-only since
  0.7.x), and document the commands that actually run here
  (`python -m unittest discover`, `ruff check`,
  `node scripts/check-version-sync.js`, `(cd rust && maturin develop)`).
- `.specs/product/VISION.md` describes the real product thesis:
  deterministic project context as a compilable artifact, not a prompt.
- `.specs/product/DOMAIN.md` lists the real entities (ProjectFile, Precedent,
  Module/Layer, Symbol, CallEdge, Endpoint, Screen, IndexState, Receipt),
  the seven critical rules (schema is contract, determinism, idempotency,
  no-LLM-on-the-critical-path, generic normalization, lightweight deps,
  cross-runtime parity) and the documented edge cases.
- `.specs/product/PERSONAS.md` captures the five real consumers (P1
  SendSprint orchestrator, P2 simplicio-dev-cli, P3 human maintainer, P4
  CI / DoD, P5 release maintainer) with the commands each calls and the
  guarantees they need.

No code changes. CLAUDE.md and .github/copilot-instructions.md stay aligned
with AGENTS.md as the repo's own contributing rule requires.

https://claude.ai/code/session_01JdmemqddwFnvbceWyuDE8m
@wesleysimplicio wesleysimplicio marked this pull request as ready for review June 2, 2026 13:34
Copilot AI review requested due to automatic review settings June 2, 2026 13:34
@wesleysimplicio wesleysimplicio merged commit 1e74e82 into main Jun 2, 2026
36 checks passed
Copy link
Copy Markdown
Contributor

Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Este PR fecha a #99 ao “dogfoodar” o próprio scaffold: substitui placeholders (ex.: <STACK>, <APP_NAME>) por contexto real do simplicio-mapper e atualiza os docs de produto em .specs/product/ para refletir tese, domínio e personas do repositório.

Changes:

  • Preenche identidade/stack/comandos reais em AGENTS.md, CLAUDE.md e .github/copilot-instructions.md.
  • Reescreve .specs/product/VISION.md, DOMAIN.md e PERSONAS.md com conteúdo específico do simplicio-mapper (artefatos, entidades, regras e consumidores).

Reviewed changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 13 comments.

Show a summary per file
File Description
CLAUDE.md Alinha instruções do Claude Code com a identidade/stack/comandos reais do repo.
AGENTS.md Atualiza contexto operacional + stack + comandos importantes do scaffold canônico.
.specs/product/VISION.md Define problema, proposta de valor e critérios de sucesso do simplicio-mapper.
.specs/product/DOMAIN.md Formaliza entidades, regras críticas e casos de borda do domínio.
.specs/product/PERSONAS.md Substitui template por personas reais (orquestrador, CLI consumidor, CI, maintainer).
.github/copilot-instructions.md Replica stack e comandos reais para o Copilot/Agent Mode.

Comment thread AGENTS.md
- Linter/formatter: `<STACK>` (sugestão: ESLint + Prettier, Ruff, dotnet format)
- CI/CD: GitHub Actions (ver `.github/workflows/`)
- Deploy: `<STACK>` (Vercel/Netlify/Docker/Azure/AWS — ver `.specs/workflow/RELEASE.md`)
- Linguagem principal: **Python 3.10+** (canonical PyPI package `simplicio-mapper`); um espelho Node 18+ vive em `bin/cli.js` + `bin/mapper-artifacts.js` mantido em paridade.
Comment thread AGENTS.md
Comment on lines +96 to +98
ruff check simplicio_mapper tests/python # Python lint
node scripts/check-version-sync.js # versões alinhadas (package/pyproject/__init__)
python -m unittest discover -s tests/python # Python unit
Comment thread CLAUDE.md
Comment on lines 27 to 28
Key placeholders to replace in real projects:

Comment thread CLAUDE.md
- Linter/formatter: `<STACK>` (sugestão: ESLint + Prettier, Ruff, dotnet format)
- CI/CD: GitHub Actions (ver `.github/workflows/`)
- Deploy: `<STACK>` (Vercel/Netlify/Docker/Azure/AWS — ver `.specs/workflow/RELEASE.md`)
- Linguagem principal: **Python 3.10+** (canonical PyPI package `simplicio-mapper`); um espelho Node 18+ vive em `bin/cli.js` + `bin/mapper-artifacts.js` mantido em paridade.
Comment thread CLAUDE.md
Comment on lines +99 to +103
npm run lint # JS + shell lint (scripts/lint.js)
ruff check simplicio_mapper tests/python # Python lint
node scripts/check-version-sync.js # versões alinhadas (package/pyproject/__init__)
python -m unittest discover -s tests/python # Python unit
node --test tests/unit # Node unit
Comment thread .specs/product/DOMAIN.md
Comment on lines +47 to +49
3. **Idempotência.** `simplicio-mapper index <path>` deve retornar exit 2
em <200 ms quando o fingerprint não mudou. Foreground e background não
podem rodar concorrentes; lock em `.simplicio/index.lock`.
Comment thread .specs/product/DOMAIN.md
```
## Casos de borda

- **Repo sem git** — `_compute_fingerprint` cai no fallback de mtimes.
Comment on lines +14 to +15
- Exit codes estáveis (0/1/2), JSON estável (`--json`), idempotência
real (<200 ms quando nada mudou).
Comment on lines +65 to +67
- Build reproduzível (`hatchling>=1.27,<1.28`).
- Versões alinhadas (`scripts/check-version-sync.js`).
- Pytest verde antes de qualquer publish.
Comment on lines +59 to +60
- **Quem** — workflows `dod.yml`, `python-ci.yml`, `scaffold-self-check.yml`,
`publish-pypi.yml`.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants