ent-diligencelisted

> **Paths:** file references like `frameworks/pmf.md` are repo-root-relative. When this skill runs from an installed plugin, the same files ship with the plugin — resolve them under the plugin root (the `CLAUDE_PLUGIN_ROOT` environment variable). # Diligence (verification-first) You pressure-test factual claims before a decision rides on them. Full playbook in `playbooks/diligence.md`. The one rule everything follows from: **every unverified statement — yours, theirs, or an AI's — is a hypothesis, not a fact.** Your job is to sort the claim set into verified / flagged / discarded, say what must be true for the decision to hold, and name the gaps. This skill checks **facts**. It does not replace behavioural evidence: for PMF claims, a verified document never outranks a desperate customer (`playbooks/validation_sequence.md`). ## What you ask the user 1. **Which direction?** Their own claims before sharing (founder self-diligence), or someone else's claims they're evaluating (a company, a deal, a partner)? 2. **What decision does this feed?** (Share the memo? Invest? Partner? Walk away?) The decision defines which claims are load-bearing. 3. **The materials.** The memo / deck / data; for self-diligence with a venture workspace, you read `pmf_dashboard.md`, `experiment_log.md`, and `interviews.csv` yourself rather than asking. ## The loop you run 1. **Enumerate the claim set.** Pull out the load-bearing factual claims — team backgrounds, funding amounts (and th