---
name: just-raised
description: Find B2B startups that raised money in the last N days, rank them by fit for what you sell, and write a first line for each one about their raise. Use when the user says "who just raised", "find startups that just got funded", "funding-triggered leads", "/just-raised", or wants a fresh outbound list of newly funded B2B companies.
---

# /just-raised

Newly funded startups have budget, a new plan and a board asking for growth. The week after a raise is the best week to
pitch them. This skill finds them, ranks them for YOUR offer, and drafts a first line for each one about their raise.

## What you need
- Claude Code with a company-data connector that can filter by last funding date. Built and tested with the Crustdata
  MCP (`crustdata_company_search_db_v2`). Any source with "last round date / type / amount" works the same way.
- Your offer in one line ("I sell outbound for B2B startups") and who it's for.

## Steps

1. **Ask** (skip what the user already said): what they sell, ideal company size, countries, stages they want
   (pre-seed / seed / Series A / later), and the window (default: last 30 days).

2. **Check the cost before pulling anything.** Data APIs bill per result, and the real bill can be far above the rate
   card. Run the search with `limit: 1` first: read `total_count` and the credits it charged. Tell the user the count and
   the estimated cost of the full pull (cost of the probe x results), and get a yes before pulling more than 25 results.

3. **Search** (Crustdata example):
   ```json
   {"filters": {"op": "and", "conditions": [
     {"field": "funding.last_fundraise_date", "type": "=>", "value": "<today minus N days, YYYY-MM-DD>"},
     {"field": "headcount.total", "type": "=<", "value": 500},
     {"op": "or", "conditions": [
       {"field": "taxonomy.categories", "type": "(.)", "value": "B2B"},
       {"field": "taxonomy.categories", "type": "(.)", "value": "SaaS"},
       {"field": "taxonomy.categories", "type": "[.]", "value": "Enterprise Software"}]}]},
    "format": "json",
    "fields": ["crustdata_company_id", "basic_info.name", "basic_info.primary_domain", "headcount.total",
               "funding.last_round_type", "funding.last_round_amount_usd", "funding.last_fundraise_date", "locations.country"]}
   ```
   Keep only venture rounds: pre_seed, seed, angel, series_a..series_e, series_unknown, convertible_note, undisclosed,
   corporate_round. Drop post-IPO, private equity, debt, grants. Spot-check the 2-3 biggest rounds with a web search
   before quoting totals.

4. **Rank by fit** (0-100). Default rubric, then adjust the weights to the user's offer and say how:
   - Stage: Series A 30, seed 26, pre-seed 14, angel 10, other 4
   - Round size: $5M-$40M 25, $2M-$5M 15, any other disclosed amount 6
   - Team size: 10-80 people 25, other known size 12
   - Country matches the user's market: 12
   - Freshness: up to 8 (8 minus one point per 4 days since the raise)
   Show a table: rank, company, round, amount, date, size, country, fit.

5. **Write the first line** for the top 10 (or as many as the user wants). One or two sentences, plain and human:
   name the raise (round + amount + when), connect it to the problem the user's offer solves, end with a soft ask.
   No hype words, no "I hope this finds you well". Example (offer: outbound for B2B startups):
   > Congrats on the $11M Series A last week. That's usually when the board wants pipeline fast. Open to a quick chat about outbound?

6. **Save** `just-raised-<date>.csv` with every column plus `fit` and `first_line`, and tell the user what it cost.

## Rules
- Never invent a company, round, amount or date. If a field is missing, leave it blank.
- Don't pull contacts (emails, phones) unless the user asks and approves the extra cost.
- Nothing gets sent. This skill drafts; the user decides.