tor-to-proposal

The bid tool that refuses to lie.

ToR + CV in → verified bid pack out — customized cover letter, tailored CV, technical response, financial proposal, submission checklist — as PDFs, in minutes, for UN / NGO / government consultants.

It does not write your evidence for you. Your track record wins bids; this tool guarantees your bid is complete, compliant, honestly priced, verified against the client's scoring grid, and audited on the final rendered bytes before you submit.

Why it exists

Every factual claim in any output must trace to a page in the source ToR, a user-supplied input, or a cited external source — otherwise it stays [FILL]. The tailored CV is proven line-by-line against your master CV. Pricing runs only on numbers you confirm. The app never fabricates rates, experience, dates, names or client facts. In a domain where one invented fact means disqualification (and vendor blacklisting), that is the feature, not a limitation.

Choose your door

Door 1 — "I use Claude / ChatGPT / Gemini" (for consultants, not coders)

You never learn the CLI. You talk to your AI; the AI runs the machinery; you answer plain questions and paste one brief into the chat per document. One-time setup (about five minutes, a couple of copy-paste commands):

Claude Code users:

npx skills add ahaomar/tor-to-proposal

Claude Desktop / other MCP apps — add the MCP server (see the Technical Guide for per-app JSON/config):

claude mcp add tor-to-proposal -- npx -y tor-to-proposal mcp-server

Any AI chat (no setup at all): run npx tor-to-proposal bid-pack start --tor tor.pdf --cv cv.pdf --dir mybid/ once with help from the User Guide, then copy out/brief-cover-letter.md and out/brief-technical-proposal.md — complete, ready-to-paste prompts — into Claude/ChatGPT/Gemini, save the replies into the bid folder, and let the tool audit the result. The AI writes; the tool refuses to let it lie.

Then, in your AI, just say: "Run a bid pack on this folder" and answer the questions it asks you — one at a time, in plain language.

Never sure what to type? The full copy-paste conversation script — every sentence to say at every stage of a bid — is in docs/CONVERSATION-SCRIPT.md.

Door 2 — "I live in the terminal" (for technical users)

npx tor-to-proposal init

That sets up your consultant profile (name, credentials, rates, master CV) at .tor-to-proposal/profile.json in the current folder (the legacy ~/.tor-to-proposal/profile.json is still read if no local profile exists).

npx skills add ahaomar/tor-to-proposal   # install as an agent skill

Local-first: no network calls carry your documents. The only optional outbound step is market-rate research you run yourself.

The workflow — ToR + CV in, bid pack out

tor-to-proposal bid-pack start  --tor tor.pdf --cv cv.txt --dir mybid/  # screen + gaps + draft CV
tor-to-proposal bid-pack apply  --answers answers.json --dir mybid/    # your answers -> pricing + financials
tor-to-proposal bid-pack pack   --dir mybid/                           # verify + checklist + zip

With an assistant, you never run these — you say "run a bid pack on this folder" and answer one consolidated questionnaire.

| Phase | What happens | You decide | | --- | --- | --- | | 1. Intake | ToR structured: deadline, effort, evaluation grid, compliance matrix, bid screen. CV gap analysis (MATCH / GAP — do not claim). Tailored CV drafted from your master CV. | Go / no-go | | 2. The one questionnaire | Every open item — pricing, availability, validity, missing evidence — in a single numbered list. apply turns answers into pricing + the financial proposal. | All of it | | 3. Drafts | Cover letter (lint-gated), technical response per scored criterion (coverage-simulated), tailored CV (anchor-verified), financial proposal (formulas visible). | Review | | 4. Pack | Every rendered PDF/docx/xlsx is re-extracted and re-audited; pack/ + submission-checklist.md + deadline.ics + zip. Stale packs are removed on failure. | Submit |

The agent-facing master workflow lives in SKILL.md. Every stage is also directly runnable: tor-to-proposal <command> --help (17 commands).

Guides

| Guide | For | | --- | --- | | docs/USER-GUIDE.md | Non-technical consultants — plain-language, step-by-step, what you'll see at every step | | docs/PROMPT-GUIDE.md | Agent users — setup for Claude/ChatGPT/Codex/any agent + a copy-paste prompt library for each phase | | docs/TECHNICAL-GUIDE.md | Developers / IT — architecture, CLI reference, data contracts, MCP/plugin internals, extension points |

Trust

• Local-first. Nothing about your bid leaves your machine.
• Honesty gates are mechanical, not aspirational: cover-letter lint, CV anchor lint, pricing gate, simulator, template-filler (client's formulas never touched), and a final audit that runs on the rendered bytes.
• pricing.md / pricing.json contain your floor math — the pack builder never includes them in what you submit.

Development

npm install
npm test     # 39 tests: extract, pricing, rates, cv-gap, cover-letter lint,
             # simulator, template-filler (docx+xlsx), audit, profile,
             # cv-tailor build/lint, financial proposal (md+xlsx), render
             # round-trip, bid-pack start/apply/pack, MCP handshake

Node ≥ 18. Dependencies (all local, all pure JS): unpdf, mammoth, exceljs, adm-zip.

License

MIT — see LICENSE. Design spec: tor-to-proposal-build-spec.md.