Configure devtronic for this project through interactive conversation. Use when setting up AI assistance in an existing project.
Interactive configuration of AI agent rules and workflows for this project.
For new projects from scratch: Use /scaffold instead - it bootstraps the project and then runs init.
This skill provides a conversational interface to the CLI. After gathering configuration, it executes:
npx devtronic init -y --ide <selected-ides>
This ensures:
Analyze the project to detect:
Read package.json to detect:
Scan directory structure to detect:
Detect technology stack:
Check for existing AI configs:
.claude/ directoryCLAUDE.md or AGENTS.md.cursor/, .agent/, etc.Present the analysis conversationally:
## Project Analysis
I've analyzed your project. Here's what I found:
**Framework**: Next.js 14.2.0
**Architecture**: Clean Architecture
**Layers detected**: domain, application, infrastructure
**Stack**:
- State: Zustand
- Data fetching: React Query
- ORM: Prisma
- Testing: Vitest
- UI: Tailwind CSS
**Scripts**:
- `npm run typecheck` - Type checking
- `npm run lint` - Linting
- `npm run test` - Tests
**Existing AI configs**: .cursor/ found
Is this correct? Would you like to adjust anything?
Ask which IDEs to configure:
Which IDEs do you use for this project?
1. **Claude Code** (current) - Will be configured
2. **Cursor** - AI-powered VS Code fork
3. **Google Antigravity** - Google's agentic IDE
4. **GitHub Copilot** - VS Code extension
You can select multiple. Claude Code will always be included since you're using it now.
If user wants adjustments:
Architecture adjustment:
User: "We actually use feature-based, not Clean Architecture"
Update configuration: Architecture = feature-based
Stack adjustment:
User: "We don't use Prisma, we use Drizzle"
Update configuration: ORM = Drizzle
Layer adjustment:
User: "Our layers are: features, shared, lib"
Update configuration: Layers = [features, shared, lib]
Once configuration is confirmed, execute the CLI:
# Build the command based on selections
npx devtronic init -y --ide claude-code,cursor
If user made adjustments that differ from auto-detection, create a temporary config or guide them to run the CLI interactively:
# For interactive mode with manual configuration
npx devtronic init --ide claude-code,cursor
After CLI execution, provide summary:
## Setup Complete!
The CLI has configured the following:
**Generated (personalized)**:
- `AGENTS.md` - With your stack (Zustand, React Query, Drizzle)
- `.claude/rules/architecture.md` - Feature-based architecture rules
**Created**:
- `.claude/skills/` - 13 workflow skills
- `.claude/agents/` - 3 specialized agents
- `.claude/rules/quality.md`
- `thoughts/` directory structure
- `CLAUDE.md -> AGENTS.md` (symlink)
**For Cursor**:
- `.cursor/rules/` - Architecture and quality rules
**Tracking**:
- `.ai-template/manifest.json` - For future updates
### Next Steps
1. Review `AGENTS.md` and customize further if needed
2. Try `/spec` to create your first feature specification
3. Add to `.gitignore`:
thoughts/checkpoints/ .claude/settings.local.json CLAUDE.local.md .ai-template/
### Future Updates
To update to newer template versions:
```bash
npx devtronic update --check # Check for updates
npx devtronic update # Apply updates
Updates preserve your local modifications.
---
## Alternative: Direct CLI Commands
For specific operations, you can also suggest:
```bash
# Add another IDE later
npx devtronic add antigravity
# Regenerate rules after stack changes
npx devtronic regenerate --rules
# Preview what would be generated
npx devtronic init --preview
Search for places (restaurants, cafes, etc.) via Google Places API proxy on localhost.
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
Create or update AgentSkills. Use when designing, structuring, or packaging skills with scripts, references, and assets.
Start voice calls via the OpenClaw voice-call plugin.
Notion API for creating and managing pages, databases, and blocks.
Gemini CLI for one-shot Q&A, summaries, and generation.
Category:developer