Creer v0.5 — content diff preview, packs, multi-root

.gitignore

@@ -0,0 +1,9 @@
+.DS_Store
+*.log
+.env
+**/venv/
+**/__pycache__/
+**/node_modules/
+**/out/
+*.pyc
+.pytest_cache/

PLAN.md

@@ -0,0 +1,38 @@
+# Creer — Final Plan
+
+v0.1 delivered the foundation: FastAPI planner/generator + VS Code command that writes a generated repo into the workspace (optional git init).
+
+## v0.2 — done
+
+1. **Preview before writing** — plan preview markdown + confirm before generate/write (`creer.previewBeforeWrite`).
+2. **GitHub repo creation** — `POST /github/create-repo` + extension remote add/push.
+3. **Curated templates** — `GET /templates` + template-anchored `/plan` & `/generate`.
+4. **Overwrite protection** — per-file conflict detection with overwrite / skip / cancel.
+5. **Chat command `/creer`** — `creer.createRepoFromChat` + `@creer` chat participant.
+
+## v0.3 — done
+
+1. **Streaming generation** — `POST /generate/stream` (SSE) + extension progress UI.
+2. **Local / offline backends** — `OPENAI_BASE_URL`, `CREER_OFFLINE`.
+3. **Open-source bake-ins** — LICENSE / README / CI via `bakeins.py`.
+4. **Hardening** — `GIT_ASKPASS` + SecretStorage for GitHub tokens.
+
+## v0.4 — done
+
+1. **Cancellation** — `job_id` on stream + `POST /generate/cancel`; extension AbortSignal + cancellable progress.
+2. **Selectable bake-ins** — license (`mit` / `apache-2.0` / `none`) and CI presets (`auto` / `python` / `node` / `none`); `GET /bakeins`.
+3. **Quality gates** — telemetry-free tree checks (`quality` on generate/done; `POST /quality`).
+
+## v0.5 (optional next)
+
+- Diff preview of generated file contents before write
+- Multi-root workspace targeting
+- Template packs as installable JSON/YAML
+
+## Non-goals
+
+- Multi-agent orchestration
+- Memory graphs
+- Overengineered plugin frameworks
+
+Stay power-focused: idea → plan → files → workspace.

README.md

@@ -0,0 +1,111 @@
+# Creer
+
+AI-powered repo scaffolding inside your workspace.
+
+**Current version: 0.5.0**
+
+## Architecture
+
+```
+creer/
+├── backend/     # Python FastAPI AI engine (+ packs/)
+└── extension/   # VS Code extension
+```
+
+## Prerequisites
+
+- Python 3.10+
+- Node.js 18+
+- OpenAI API key **or** `OPENAI_BASE_URL` **or** `CREER_OFFLINE=1` (with template/pack)
+- VS Code / Cursor
+
+## Backend (v0.5)
+
+```bash
+cd backend
+python -m venv venv && source venv/bin/activate
+pip install -r requirements.txt
+cp .env.example .env
+uvicorn main:app --reload --port 8000
+```
+
+### Environment
+
+| Variable | Description |
+|---|---|
+| `OPENAI_API_KEY` | OpenAI API key |
+| `OPENAI_BASE_URL` | OpenAI-compatible base URL (Ollama, etc.) |
+| `CREER_MODEL` | Model name (default `gpt-4o-mini`) |
+| `CREER_OFFLINE` | Template/pack-only stubs |
+| `CREER_PACKS_DIR` | Extra packs directory (overrides same ids) |
+
+### Endpoints
+
+| Method | Path | Description |
+|---|---|---|
+| `GET` | `/health` | Version `0.5.0`, packs_count |
+| `GET` | `/templates` | Built-in templates |
+| `GET` | `/packs` | Installable JSON/YAML packs |
+| `GET` | `/packs/{id}` | Single pack |
+| `GET` | `/bakeins` | License/CI options |
+| `POST` | `/plan` | Plan (`template_id` **or** `pack_id`) |
+| `POST` | `/generate` | Generate + bake-ins + quality |
+| `POST` | `/generate/stream` | SSE progress (`job_id`) |
+| `POST` | `/generate/cancel` | Cancel job |
+| `POST` | `/quality` | Dry-run gates |
+| `POST` | `/github/create-repo` | Create GitHub repo |
+
+### Packs
+
+Drop `.json` / `.yaml` files into `backend/packs/` (or `CREER_PACKS_DIR`):
+
+```json
+{
+  "id": "fastapi-crud",
+  "name": "FastAPI CRUD",
+  "description": "CRUD API starter",
+  "stack": "FastAPI + Uvicorn",
+  "version": "1.0.0",
+  "files": ["main.py", "requirements.txt", "README.md"]
+}
+```
+
+Shipped examples: `fastapi-crud`, `express-ts`, `python-lib`.
+
+## Extension (v0.5)
+
+```bash
+cd extension && npm install && npm run compile
+```
+
+Flow: idea → template/pack → bake-ins → plan preview → generate (cancellable) → **content diff preview** → write → optional git/GitHub.
+
+Multi-root: QuickPick workspace folder (or `creer.defaultWorkspaceFolder`).
+
+### Commands
+
+| Command | Title |
+|---|---|
+| `creer.createRepo` | Creer: Create New Repo |
+| `creer.createRepoFromChat` | Creer: Create from Chat Prompt |
+| `creer.setGitHubToken` | Creer: Set GitHub Token |
+| `creer.clearGitHubToken` | Creer: Clear GitHub Token |
+
+### Settings
+
+| Setting | Default | Description |
+|---|---|---|
+| `creer.backendUrl` | `http://localhost:8000` | Backend URL |
+| `creer.contentPreview` | `true` | Diff/content preview before write |
+| `creer.defaultWorkspaceFolder` | `""` | Multi-root folder name/path hint |
+| `creer.previewBeforeWrite` | `true` | Plan tree confirm before generate |
+| `creer.useStreaming` | `true` | SSE progress |
+| `creer.promptBakeins` | `true` | QuickPick license/CI |
+| `creer.license` / `creer.ciPreset` | `mit` / `auto` | Defaults when not prompting |
+| `creer.initGit` | `true` | git init + commit |
+| `creer.createGitHubRepo` | `false` | Create remote |
+| `creer.githubPrivate` | `true` | Private repos |
+
+## License
+
+MIT

backend/.env.example

@@ -0,0 +1,17 @@
+# OpenAI API key (required unless CREER_OFFLINE=1 or using a local server that ignores it)
+OPENAI_API_KEY=sk-your-key-here
+
+# Optional OpenAI-compatible base URL for local/offline backends (Ollama, LM Studio, vLLM, etc.)
+# Example: http://127.0.0.1:11434/v1
+OPENAI_BASE_URL=
+
+# Model name (OpenAI or local-compatible)
+CREER_MODEL=gpt-4o-mini
+
+# Template-only / stub generation — never calls the LLM
+# Set to 1, true, or yes to enable
+CREER_OFFLINE=
+
+# Optional extra packs directory (JSON/YAML). Merged with backend/packs;
+# user packs override built-in packs on the same id.
+# CREER_PACKS_DIR=/path/to/my-packs

backend/.gitignore

@@ -0,0 +1,6 @@
+.env
+venv/
+__pycache__/
+*.pyc
+.pytest_cache/
+.mypy_cache/

backend/app/__init__.py

@@ -0,0 +1 @@
+"""Creer backend package."""