agentleFS
mardab96mardab96/b2b-lead-generation-claude-skills/AGENTS.md

b2b-lead-generation-claude-skills

AGENTS.md5 starsChanged 4 months agoMITView on GitHub

A pack of 30 Claude Skills for B2B lead generation. Each skill is a self-contained SKILL.md in its own folder, following the Agent Skills standard, so the pack runs in any Agent Skills tool, not only Claude Code. - Evidence first. Every skill works from what you supply: a CRM export, a transcript, an email thread, a DNS record, a company website. Where evidence is missing the skill marks the finding as an assumption rather than filling the gap with…

The file

We split this AGENTS.md into the rules it gives an agent

7 rules. About 0 are common, 7 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    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:

    Evidence first: Every skill works from what you supply: a CRM export, a transcript, an email thread, a DNS record, a company website.

    line 9
  2. Distinctive:

    Nothing goes out: No skill sends an email, edits a CRM record, changes a sequence or touches DNS.

    line 13
  3. Distinctive:

    No invented proof: No fabricated customer quotes, trigger events, funding rounds, case study numbers or reply-rate promises.

    line 16
  4. Distinctive:

    Do not run a skill without its required input.

    line 76
  5. Distinctive:

    Do not present an inference about a prospect's internal situation as fact.

    line 77
  6. Distinctive:

    Do not treat one ghosted deal, one bad call or one lost deal as a pattern.

    line 78
  7. Distinctive:

    Do not send, publish or apply anything a skill produced without a human reading it first.

    line 79

≈ “Common” is an estimate: at least 1 in 5 of the 41,581 AGENTS.md files we read have a rule of the same kind, not the same words. “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
AGENTS.md

AGENTS.md: B2B lead generation skills

A pack of 30 Claude Skills for B2B lead generation. Each skill is a self-contained SKILL.md in its own folder, following the Agent Skills standard, so the pack runs in any Agent Skills tool, not only Claude Code.

How these skills behave

  • Evidence first. Every skill works from what you supply: a CRM export, a transcript, an email thread, a DNS record, a company website. Where evidence is missing the skill marks the finding as an assumption rather than filling the gap with a plausible number.
  • Nothing goes out. No skill sends an email, edits a CRM record, changes a sequence or touches DNS. They produce text and recommendations that a human sends or applies.
  • No invented proof. No fabricated customer quotes, trigger events, funding rounds, case study numbers or reply-rate promises. This matters more in B2B than elsewhere, because the buyer usually knows the truth about their own company and notices immediately.

How the pack splits

Skills 1 to 20 are the inbound and paid machine: targeting, pages, forms, scoring, source quality, handoff. They assume leads arrive and ask whether the right ones are arriving.

Skills 21 to 30 are the outbound week and the sales side. They assume nothing arrives and ask what you send, what you say on the call, and what you do with the silence afterwards.

The two halves answer different questions and are rarely useful at the same moment. A team with plenty of leads and a quality problem lives in the first twenty. A team with an empty pipeline lives in the second ten.

Order that matters

  • spam-folder-check runs before cold-outbound-sequence-review, always. A delivery problem and a copy problem produce the same symptom, and rewriting copy fixes only one of them.
  • ghosted-after-the-demo handles one silent deal. When the same silence appears across many deals at the same stage, it stops being a message problem and discovery-call-gap-analysis takes over.
  • pipeline-hygiene-audit comes before any forecast conversation, not after one goes wrong.
  • objection-cheat-sheet and won-deal-to-case-study both read historical deals, so they are cheap to run together on the same export.

Where judgement shifts

  • Deal size. Under a few thousand, most of the second ten are too heavy; the volume game is the first twenty. Above five figures, per-account work like what-to-say-to-this-company and meeting-prep-five-minutes pays for itself on a single meeting.
  • Sales cycle length. A long cycle makes pipeline-hygiene-audit essential and makes stale-deal thresholds much more forgiving. Set --cycle-days rather than accepting a default.
  • Who does the selling. A founder selling directly needs the prep and follow-up skills. A team with SDRs needs the sequence, transcript and pipeline skills, because the failure is consistency rather than capacity.
  • Regulated markets. Cold outreach rules differ by jurisdiction. No skill in this pack gives legal advice, and several say so explicitly.

Scripts

Two skills carry deterministic helpers, because their work is counting and dates rather than judgement: pipeline_hygiene.py and email_auth_check.py. selftest.py locks what they guarantee, including the defect where a future close date was allowed to anchor the reference date and marked healthy deals as overdue. Run it before and after changing anything in scripts/.

The other eight skills read text and produce text. A script there would add ceremony, not rigour.

What not to do

  • Do not run a skill without its required input. Skip it and say what is missing.
  • Do not present an inference about a prospect's internal situation as fact.
  • Do not treat one ghosted deal, one bad call or one lost deal as a pattern.
  • Do not send, publish or apply anything a skill produced without a human reading it first.

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.