lead-research-dossier
Skill4 starsChanged 4 months agoMITView on GitHub
Build a structured one-page prospect dossier (company facts, ICP fit, buying signals, contacts, personalization hooks) from live web research. Use when the user asks to "research this lead", "build a dossier", "who is this company", "prep me for this sales call", "qualify this prospect", or names a company they want to sell to. Don't use for writing the outreach itself — hand the finished dossier to cold-outreach-sequencer.
In this repository74 other files it gives its agents.
- agent-eval-harnessSkill
- context-budget-managerSkill
- incident-rollbackSkill
- memory-hygieneSkill
- model-cost-optimizerSkill
- prompt-debuggerSkill
- skill-author-coachSkill
- third-party-skill-vetterSkill
- api-endpoint-testerSkill
- bug-triage-protocolSkill
- changelog-release-notesSkill
- docs-readme-writerSkill
- env-secrets-hygieneSkill
- github-pr-reviewSkill
- git-ship-flowSkill
- lovable-brief-writerSkill
- supabase-migration-writerSkill
- vercel-deploy-checkSkill
- cold-outreach-sequencerSkill
- competitor-pricing-scanSkill
- contract-redline-scanSkill
- customer-support-inboxSkill
- invoice-runnerSkill
- kpi-dashboard-weeklySkill
- payment-reconciliationSkill
- pipeline-trackerSkill
- proposal-builderSkill
- stripe-payment-linkSkill
- testimonial-harvesterSkill
- weekly-revenue-reportSkill
- brand-voice-guardianSkill
- case-study-builderSkill
- content-repurposerSkill
- deep-research-briefSkill
- email-nurture-sequenceSkill
- landing-page-copywriterSkill
- linkedin-content-packSkill
- newsletter-assemblerSkill
- seo-blog-writerSkill
- social-listening-reportSkill
- x-thread-builderSkill
- youtube-script-writerSkill
- budget-guardrailsSkill
- ceo-operating-doctrineSkill
- delegation-protocolSkill
- editor-personaSkill
- heartbeat-standupSkill
- hiring-briefSkill
- ops-agent-personaSkill
- researcher-personaSkill
- sales-agent-personaSkill
- support-agent-personaSkill
- weekly-retroSkill
- writer-personaSkill
- calendar-conciergeSkill
- decision-journalSkill
- email-triageSkill
- file-organizerSkill
- habit-trackerSkill
- inbox-zero-sprintSkill
- The first 60 shown.
The file
Files
Lead Research Dossier
Produces a one-page, evidence-backed prospect dossier an SDR could act on in five minutes. Every claim carries a source URL; anything unverifiable is marked "unconfirmed" rather than guessed. The dossier feeds directly into cold-outreach-sequencer.
When to Use
- User names a company (or a person + company) they want to sell to, partner with, or pitch.
- User asks for call prep, account research, or "is this company a fit for us?"
- User has a list of leads and wants dossiers one at a time.
- Not for: writing the outreach itself — hand the finished dossier to
cold-outreach-sequencer. Not for competitor analysis — usecompetitor-pricing-scan.
Quick Reference
| Action | Command / Call |
| Company site sweep | web_extract on homepage, /about, /pricing, /careers, /blog |
| Recent signals | web search: "<company>" (funding OR hiring OR launch OR partnership) 2026 |
| Team + contacts | web search: "<company>" (founder OR "head of" OR VP) site:linkedin.com |
| Tech/scale hints | careers page job posts (stack, team size), press releases |
| ICP fit scoring | Compare findings against the user's ICP criteria (ask if not given) |
Procedure
- Precheck — confirm web tools are available and you have: company name or domain, and the user's ICP criteria (target size, industry, problem you solve, who buys). If ICP criteria are missing, ask for them or pull from prior context — fit scoring is meaningless without them.
- Firmographics — extract homepage + about page: what they do, for whom, HQ, rough headcount, business model. Success: you can state their one-liner in your own words with a source URL.
- Signals sweep — search for events in the last ~12 months: funding, leadership changes, product launches, hiring spikes, expansion. Each signal gets a date and URL. Old news labeled as old.
- Contacts — identify 1–3 likely buyer-persona people (name, role, public profile URL). Public professional info only — no personal emails, no scraped private data. If no contact is findable, say so; never invent names.
- Hooks — from their blog, job posts, founder posts, or product pages, pull 2–3 specific, recent, non-generic personalization hooks ("they just opened a Singapore office", not "I love your website").
- Score fit — rate ICP fit High / Medium / Low against each of the user's criteria, one line of evidence per criterion.
- Deliver — fill the template completely. Mark any empty section "none found" — a visible gap beats silent omission.
Output Template
# Dossier: <Company> — <date>
**One-liner:** what they do, for whom. [source]
**Firmographics:** HQ | ~headcount | model (B2B SaaS / services / ...) | est. stage/revenue (unconfirmed if inferred)
## ICP Fit: High / Medium / Low
| Criterion | Verdict | Evidence |
|---|---|---|
## Signals (last 12 months)
- <date> — <signal> [source]
## Contacts
- <Name> — <Role> — <public profile URL>
## Personalization Hooks
1. <specific, recent, sourced>
2. ...
## Open Questions / Risks
- <what we couldn't verify>Pitfalls
- Confidently wrong facts (stale or hallucinated) — headcount, funding, and even the product change fast. Recovery: only state what a fetched page actually says, attach the URL, and date every signal. If two sources disagree, show both.
- Wrong company (name collision) — "Mercury" the bank vs. Mercury Marine. Recovery: lock onto the exact domain early; verify industry + HQ match the user's intent before deep research.
- Generic hooks — "congrats on your growth" gets deleted on sight. Recovery: a hook must pass the test "could this sentence apply to any other company?" If yes, dig into their blog/job posts until you find one that couldn't.
- Empty contacts section treated as blocker — some companies expose no people publicly. Recovery: deliver the dossier anyway with contacts marked "none found publicly" and suggest the role title to target instead.
Verification
- Every factual claim has a source URL; inferences are labeled "unconfirmed"
- All signals dated and from the last 12 months (or flagged as older)
- Hooks are specific enough that they couldn't apply to another company
- ICP fit scored against the user's actual criteria, not generic ones
- No private/personal contact data included — public professional info only
- Output matches the template with no silently omitted sections
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
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.

