agentleFS
iangogenticiangogentic/venture-lab/.claude/skills/harvest-leads/SKILL.md

harvest-leads

Skill0 starsChanged 4 months agoMITView on GitHub

Turn clustered evidence into leads - people, quotes, permalinks.

The file

Files
SKILL.md

Harvest leads

Each pain cluster below comes with candidates: evidence items that carry an author and a permalink. Your job is to decide which candidates show a person actually expressing the cluster's pain, quote the passage that shows it, and read what their words express.

You are selecting, not writing. Every quote you keep is checked against the excerpt it came from; anything not literally present there is rejected and the whole reply fails. You cannot add a person, a URL, or a candidate that is not below.

The pain clusters

${clusters}

The candidates, per cluster

${candidates}

Observation before inference

The quote is the observation: the person's exact words. The intent is your inference from those words and nothing else — not from the cluster's label, not from what a person probably meant. If the quote does not support the intent on its own, pick a different quote or skip the candidate.

Intent

  • seeking — actively asking for a solution. "What are people using instead of X? I can't keep doing this manually."
  • complaining — expressing the pain unprompted: venting, describing the cost, nobody asked. "We lose half a day every release to this sync step."
  • mentioning — the topic is named but no pain is expressed. "We use X for our deployments alongside Y."

Give intent_rationale in one line: which words carried the reading.

Quotes are copied, not composed

Copy the passage exactly as it appears in that candidate's excerpt. Reflowing across lines is fine; changing words, tidying, cutting mid-sentence or stitching two passages together is not. Pick the passage a reader could see alone and understand what the pain is.

evidence_id and cluster_id must match a candidate entry above exactly.

Skipping is correct

A candidate that names the topic without expressing pain, or whose excerpt is about someone else's problem, is not a lead — leave it out. A cluster yielding zero leads is a true result, not a failure. Padding the list is the failure mode here.

Never invent contact details

The permalink is the only channel. Do not infer, guess, or construct emails, handles on other platforms, employers, or real names — not even when the text hints at them. Engaging with a lead happens later, as a public, human-written reply on the platform where the person chose to speak; nothing you produce here automates or addresses that contact.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.