@bilig/create-workpaper

@bilig/create-workpaper

Create a runnable Bilig WorkPaper starter for Node services and tool integrations.

npm create @bilig/workpaper@latest pricing-workpaper
cd pricing-workpaper
npm install
npm run smoke

The generated starter builds a quote-approval workbook with the A1 facade, writes quote inputs through one atomic editManyAndReadback proof, recalculates formulas, persists JSON, restores the workbook, and prints verified: true. Generated projects pin @bilig/workpaper to the generator package version and use exact dev-tool versions instead of latest, so the smoke proof is reproducible.

After the smoke proof passes, keep the JSON output limited to proof fields. If the workflow is relevant, watch releases for API and formula compatibility updates: https://github.com/proompteng/bilig/subscription. If it almost works, open one concrete blocker or integration note: https://github.com/proompteng/bilig/discussions/new?category=general.

Use this when you want to evaluate @bilig/workpaper from a blank directory without cloning the full monorepo.

For agent and MCP evaluation, generate the agent-ready starter:

npm create @bilig/workpaper@latest pricing-agent -- --agent
cd pricing-agent
npm install
npm run agent:verify
npm run mcp:server

The agent template adds AGENTS.md, CLAUDE.md, GEMINI.md, a Claude Code project skill and proof command, GitHub Copilot / VS Code instructions, Cursor, Kiro, Roo Code, Trae, Qodo setup docs, Zed, Junie, Aider conventions, Cline, Continue, Cascade/Devin, Windsurf rules, an OpenHands skill at .agents/skills/bilig-workpaper/SKILL.md, OpenCode files at opencode.jsonc and .opencode/agents/bilig-workpaper.md, a project-root .mcp.json for Claude Code, Cursor, Kiro, Roo Code, Trae (.trae/mcp.json), Qodo IDE Agentic Tools, Zed (.zed/settings.json), Junie, Continue, and VS Code MCP configs, a generic MCP config at mcp/bilig-workpaper.mcp.json, and an agent:verify script that runs the service smoke plus the package-owned basic and revenue-plan agent evaluator proofs. The revenue-plan evaluator checks MCP tool discovery, mutation, recalculated SUM, SUMIF, XLOOKUP, FILTER, a named expression, persistence, and restart readback. The raw MCP challenge remains available as npm run mcp:challenge when you need the lower-level JSON-RPC transcript.

To add the same agent and MCP proof loop to an existing Node repo without replacing its app, run:

npm create @bilig/workpaper@latest . -- --add-agent
npm exec --yes --package @bilig/workpaper@latest -- bilig-evaluate --door agent-mcp --scenario revenue-plan --json
npm exec --yes --package @bilig/workpaper@latest -- bilig-workpaper-mcp --workpaper ./.bilig/pricing.workpaper.json --init-demo-workpaper --writable

That writes Bilig-specific agent files, keeps an existing README.md, does not edit package.json, keeps WorkPaper state under ./.bilig/pricing.workpaper.json, and skips existing files unless you pass --force. When existing agent instruction files block part of the install, the CLI writes BILIG_WORKPAPER_INSTALL.md with the skipped paths and a short handoff snippet you can paste into your current agent policy. The overlay uses your existing package.json name for BILIG_WORKPAPER.md instead of rendering the directory argument.

For an existing agent or MCP client that does not need a generated project yet, use the handoff checklist first: https://proompteng.github.io/bilig/agent-adoption-kit.html.