Turn scattered user feedback into insights you can act on
Turn interviews and feedback into structured, prioritized, evidence-backed insights.
Early-stage VCs drown in inbound. Two free Agent Skills turn deal flow into structured screens and thoughtful founder responses — a one-page pass/consider/deep-dive card, then a personalized reply that preserves the relationship. Learn the frameworks and run them in Claude.
Free to try. One click loads the skill into Claude or ChatGPT, with no account and no setup.
Early-stage investing is a volume game with a reputation problem baked in. A partner might see 50–100 inbound pitches a week; the vast majority are a pass. The volume makes two things fall apart: consistency (screening quality drifts by mood and time of day) and courtesy (founders get ghosted, which quietly damages the fund’s brand and its referral network).
Doing it well means separating those jobs and systematizing both: a fast, structured screen so good deals aren’t missed and bad ones aren’t over-analyzed, and a thoughtful reply to every founder so the relationship survives the “no.” That’s exactly what these two skills encode — the screening rigor of a good associate and the relationship instincts of a good partner, made repeatable.
The skill isn’t magic. It encodes proven frameworks anyone can learn. Here’s the method itself, so you understand what it’s doing for you.
The problem for any early-stage investor is triage: 50-plus inbound a week, and only so many hours. The framework splits screening into two tiers so effort matches promise. The Quick Screen Card is a one-page pass / consider / deep-dive recommendation generated from a single input (a deck, a cold email, an intro blurb, or one URL) — it answers “should I spend time on this?” in seconds. Only deals that clear the card graduate to the Full Deal Assessment: an 8–15 page evaluation built through classify → research → assess → synthesize, ready for the investment committee.
The discipline is that the card is always generated first, and the full assessment builds on it. That keeps you from sinking hours into deals that a one-pager would have screened out — and gives every promising deal a consistent, structured evaluation instead of a gut call.
Every screen produces a verdict, and every verdict deserves a reply — because a founder you pass on today is a referral source, a repeat founder, or a future deal. The framework maps the verdict to one of three response types. A Warm Pass (verdict: PASS) — a kind, specific, honest decline that leaves the door open. A Request for Info (verdict: CONSIDER) — a targeted follow-up that keeps the deal alive while you fill the gaps. A Meeting Request (verdict: DEEP DIVE) — a warm, genuine invite that moves fast on a promising deal.
Each is tuned to the relationship context too (cold inbound vs. warm intro vs. portfolio referral). The payoff: you reply to everyone, thoughtfully, without it eating your week — which is how reputations and deal flow compound.
Use any one on its own, or chain them. Each hands its output to the next. Every one is free in the public library.
Turns an inbound deck, cold email, or intro into a structured screen — a one-page pass/consider/deep-dive card, or a full 8–15 page IC assessment when it earns one.
Drafts the reply the screen calls for — a warm pass, a request for missing info, or a meeting invite — personalized to preserve the founder and referral relationships.
Any time inbound outpaces your time to evaluate it thoughtfully — which is most early-stage investors, most weeks. Especially useful for solo GPs, angels, and small teams without an associate bench to triage the pile.
Together they turn the inbound firehose into a tidy pipeline. The screener reads whatever the founder sent — a deck, a cold email, an intro — and produces a structured screen: a one-page card for fast triage, or a full IC-ready assessment for the deals worth it. The responder then drafts the reply that screen calls for, personalized to the verdict and the relationship.
The result is a repeatable deal-flow motion: every inbound gets screened consistently, and every founder gets a thoughtful, relationship-preserving reply — the two things that are hardest to do at volume and easiest to let slide.

Real prompts, and what the skill hands back.
“Here’s a cold inbound deck for a seed-stage fintech. Give me a quick screen card with a pass/consider/deep-dive recommendation.”
A one-page screen: deal archetype, key strengths and concerns, and a pass / consider / deep-dive verdict with rationale — enough to decide whether to spend more time.
“We’re passing on this one, but it came through a portfolio founder. Draft a warm pass that keeps both relationships intact.”
A kind, specific, honest decline tuned to the portfolio-referral context — respectful of the founder and the referral source, leaving the door open.
A one-page screen decides whether a deal earns deeper work — so you never over-analyze a pass.
Only deals that clear the card get the full 8–15 page IC assessment.
Every verdict maps to a response type, so no founder gets ghosted.
Replies are tuned to context — cold inbound, warm intro, portfolio referral — not templated.
Customize screening criteria and reply voice so it sounds like your fund.
It’s a clean two-step pipeline: screen → respond. vc-deal-screener produces the verdict; vc-founder-responder reads that verdict and drafts the matching reply. Run the screener alone to triage, or the pair to close the loop on every inbound deal.
Run both skills in Claude and tune them to your thesis and voice:
Already connected your library to Claude? Tell it: “Use the vc-deal-screener skill (and the others above) from the agentman_skills MCP server.”
Not connected yet? The Try in Claude / ChatGPT button loads the first one from the public library, with no setup. Or connect your library.
Click “Try in Claude / ChatGPT” above, or — if you’ve connected your library — tell Claude: “Use the vc-deal-screener skill from the agentman_skills MCP server.” Add vc-founder-responder by slug for the replies.
Give vc-deal-screener the deck, cold email, or intro. It classifies the deal archetype and returns a Quick Screen Card with a pass / consider / deep-dive verdict.
For deals that clear the card, ask for the Full Deal Assessment — it researches the company, founders, market, and competitors into an IC-ready document.
Hand the verdict to vc-founder-responder — it drafts the warm pass, request-for-info, or meeting invite that matches, tuned to the relationship context.
Customize the screening criteria to your thesis and the response tone to your voice, so every screen and reply sounds like your fund.
That’s the work of an associate triaging the pile plus a partner keeping relationships warm — from two free skills you tune to your fund.
One click loads the skill into Claude or ChatGPT. No account, no setup.
No — it structures the decision, it doesn’t make it. The screener gives you a consistent, evidenced pass/consider/deep-dive recommendation and, when warranted, an IC-ready assessment. The judgment stays yours; the skill removes the triage grind and inconsistency.
No. Run vc-deal-screener alone to triage your pipeline, or add vc-founder-responder to also draft the founder replies. Together they close the loop from inbound deck to sent response.
Yes. Clone the skills into an Agentman workspace and customize the screening criteria to your thesis and the response tone to your voice, so every screen and reply reflects how your fund operates.
No. Click “Try in Claude / ChatGPT” above and the screener loads from our public library — no setup. Clone the skills to customize and share them.