agentleFS
matteotittamatteotitta/genesys-skills/skills/icp-research/SKILL.md

icp-research

Skill53 starsChanged 5 days agoMITView on GitHub

Builds an ideal customer profile from a company's own website: case studies, testimonials and solutions pages become personas, segments, pains and buying triggers, each backed by evidence. Use when asked for ICP research, customer personas or who a company sells to.

5 other projects use a version of this, 1 of them edited.

Adopted by

Most recently taken up by

Used with

Projects with it also use win-loss, level and 2 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • calven-aiSkill

    win-loss

    Synthesise why deals were won and lost this period from closed deals and call transcripts, by competitor, segment and reason. Use when "why did we lose Acme", "quarterly win/loss".

    Also in 3 projects that use this

  • dachi-jp3Skill

    level

    表示レベル(auto / low / medium / high)を切り替える。

    Also in 3 projects that use this

  • VercelSkill

    messaging

    Use when turning settled positioning into the lines other people will reuse: the message hierarchy, value propositions per audience, objection handling, and the proof behind each claim.

    Also in 4 projects that use this

  • ai-agents-for-agencies-coachesSkill

    brand-kit

    Establish and enforce per-client brand fidelity using the DESIGN.md standard (google-labs-code/design.md). Use whenever starting work for a client, creating any ad/landing-page/creative, or when brand colors/fonts/spacing are needed. Auto-extracts a client's visual identity into a validated DESIGN.md, keeps a central registry, and feeds the tokens into image-studio (statics), fb-ad-video-studio (video), and landing pages. Triggers: "set up brand for <client>", "make a DESIGN.md", "match the client's brand", "what are their brand colors", "new client onboarding", any creative work that must be on-brand.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

2 rules. 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Never invent customer names: Only cite publicly referenced customers.

    line 81
  2. Distinctive:

    Quote verbatim: Use exact customer language in quotes.

    line 82

“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

ICP research skill

Generate ideal customer profiles for B2B SaaS clients through systematic research and structured output.

Report structure

The final ICP report follows this numbered section order:

SectionPurpose
HeaderResearch date, website, category, confidence score (1-5)
1. Executive summaryHigh-level synthesis of findings and strategic recommendations
2. TAM analysisMarket sizing with targeting strategy per layer (TAM/SAM/SOM/ICP)
3. Firmographics analysisGeographic, industry, company segment patterns, and technographics
4. Roles and personasCore use case, Champion deep-dive, Economic Buyer deep-dive, buying journey
5. Negative ICPWho is NOT a fit, disqualification criteria, and red flags
6. Customer proof pointsNamed customers, outcomes, and evidence with URLs
7. Voice of customer synthesisLanguage patterns, pain points, and outcome terminology
8. ICP segment definitionsScoring matrix, in-market signals, segment deep-dives
9. Intent signals and buying triggersObservable signals indicating purchase readiness
10. RecommendationsPrioritization and messaging by segment
11. Data gapsMissing information and follow-up suggestions
12. Source appendixAll sources with access dates, URLs, and confidence levels

Confidence score calculation: Count High/Medium/Low data points. Score 5 if >70% High, Score 4 if >50% High, Score 3 if mixed, Score 2 if >50% Low, Score 1 if >70% Low.


Sorting rules

Apply consistently across all tables:

DimensionSort order
Decision roleChampion → Economic Buyer → User → Influencer
Company sizeEnterprise → Mid-market → SMB → Startup
FrequencyVery high → High → Medium → Low
ConfidenceHigh → Medium → Low
Customer concentrationHigh → Medium → Low
Priority1 → 2 → 3 → 4
Industry presenceStrong → Moderate → Emerging

Workflow

The research runs in 3 phases. Read the premium reference for the full step-by-step.

Phase summary:

  1. Data extraction — discover key pages (customers, case studies, solutions, pricing, integrations, G2, LinkedIn) → extract raw data with URL+date per source → normalize attributes (geography, industry, company size, team size, tech stack)
  2. Analysis and synthesis — identify patterns per segment → build Champion + Economic Buyer deep-dives → identify negative ICP + intent signals → collect proof points → document technographics → calculate TAM with targeting strategy → identify ICP as highest-priority segment below SOM
  3. Structured output — generate the 12 numbered sections, apply sorting rules, include rich descriptions with URLs and dates

Input requirements

Required

  • Website URL — primary company website

Optional (improves quality)

InputPurpose
Case studies URLDirect link to case studies page
Testimonials URLDirect link to testimonials
Market contextCategory, competitors, GTM approach
Sales call notesWin/loss context, objections
Existing ICP docsValidate or expand current understanding

Anti-hallucination guardrails

  1. Never invent customer names. Only cite publicly referenced customers.
  2. Quote verbatim. Use exact customer language in quotes.
  3. Mark confidence levels. Tag data as High/Medium/Low confidence.
  4. Cite sources with URLs and dates. Include URL and access date for every claim.
  5. Acknowledge gaps. Explicit "Not available" for missing data.
ConfidenceDefinition
HighDirect from official source, verifiable
MediumThird-party source, multiple signals
LowSingle indirect source, inferred

Quality

Pre-delivery checklist (coverage / personas / segments / evidence): the premium reference.

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.