AI Automatic Proposal Generator Flow

AI Automatic Proposal Generator Flow

Before you run this

By default this skill uses local CSV files for its data (input and output) — nothing to connect, works offline, your data stays on your machine. The CSVs live next to the skill in ./data/ (input: data/input.csv, output: data/output.csv), or point it at any path you like.

If you'd rather read/write a Google Sheet instead, just say so and I'll switch you over. It's a one-time setup: you'll paste a Google service-account JSON (or authorize once), share your Sheet with that account's email, and give me the Sheet URL. After that it behaves exactly the same, just backed by your Sheet. Say "use Google Sheets" to start that, or "keep CSVs" (default) to just go.

What it does

Takes post-discovery-call notes (filled in via Typeform or manually into CSV) and auto-generates a complete sales proposal — problem pitch, solution pitch, 4 milestones with timelines, and pricing — then creates a signed PandaDoc document from a template. The original Make.com flow was: Typeform webhook trigger → fetch latest response → filter for "Send proposal" outcome → GPT-4 generates proposal JSON → GPT-4 generates milestones JSON → PandaDoc creates document from template.

This skill replaces the webhook/polling loop with a simple CSV-driven CLI you can run any time. Each row in data/input.csv is one discovery call. Rows where outcome != "Send proposal" are skipped.

When to trigger

• After a discovery call where you want to auto-generate and send a proposal
• Batch: drop multiple rows in data/input.csv, run once, get all proposals created in PandaDoc
• Testing: populate data/input.csv with fake data, set DRY_RUN=1 to skip PandaDoc and just write proposal JSON to data/output.csv

Required env vars

OPENAI_API_KEY=sk-...
PANDADOC_API_KEY=...
PANDADOC_TEMPLATE_ID=dbmxWR7jB6y3E6TzdC2buT   # "[Company Name] LeftClick Proposal"
PANDADOC_RECIPIENT_ROLE=Client                  # must match the template's role name exactly
SENDER_FIRST_NAME=Nick
SENDER_LAST_NAME=Saraev
SENDER_EMAIL=info@leftclick.ai
SENDER_COMPANY=LeftClick
DRY_RUN=0   # set to 1 to skip PandaDoc and write output.csv only

Step-by-step procedure

1. Populate input — fill data/input.csv with one row per discovery call. See the sample file. Required columns: client_first_name, client_last_name, client_email, client_company, business_description, problem, solution, tools, timeline, project_price, platform_costs, outcome. Only rows where outcome == "Send proposal" are processed.

2. Run the pipeline:

   cd /path/to/skill/
   python3 scripts/run_pipeline.py --input data/input.csv --output data/output.csv
   

   Optional flags:

   • --dry-run — generate proposal JSON, write to output CSV, skip PandaDoc
   • --input PATH — override input CSV path
   • --output PATH — override output CSV path

3. What happens per row:

   • Skips rows where outcome != "Send proposal"
   • Calls scripts/generate_proposal.py → GPT-4 returns {problemPitch, solutionPitch, title, platformList}
   • Calls scripts/generate_milestones.py → GPT-4 returns {milestone-1..4, timeline-1..4}
   • Calls scripts/create_pandadoc.py → creates PandaDoc document from template, populates all tokens and pricing tables
   • Writes result row (doc ID, doc URL, status) to output CSV

4. Check output — data/output.csv has one row per processed lead with pandadoc_doc_id, pandadoc_doc_url, proposal_title, status.

Scripts

Node map (original Make.com → this skill)