agentleFS
blitzcrieg1blitzcrieg1/agentmetry/.claude/skills/outreach-next/SKILL.md

outreach-next

Skill40 starsChanged 35 days agoApache-2.0View on GitHub

Report the next Agentmetry outreach message to send and its exact text. Use when the user asks what commercial work is next, about the pipeline, design partners, sales, or which opener to send.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not reorder this without saying why.

    line 61
  2. Very specific:

    Update the row in the private outreach-log.md with the date and channel.

    line 66
  3. Distinctive:

    It will not send anything: No email or messaging tool is available in this session, and outreach on someone's behalf is theirs to press …

    line 72

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill 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
SKILL.md

Outreach next

This is the department that has shipped nothing. Treat it accordingly.

Read, do not invent

../agentmetry-enterprise/sales/outreach-log.md  the queue, recipients, status, replies (private repo)
docs/commercial/outreach-openers.md              the six openers and the sending order (public)

The log holds named recipients with real contact details, which is why it lives in the private repo. Never copy a name, an address or a reply into this public repo, including into this skill. Do not invent a recipient, do not substitute a different company, and do not rewrite an opener into something more enthusiastic. Each opener is built around one checkable fact and states a limitation before asking for anything. That structure is the reason it works.

Report

  1. The next unsent row in the queue: opener number, person, channel
  2. Its exact text, ready to paste
  3. Anything in it that needs verifying before it goes out
  4. The current sent/reply count, honestly. It has been zero for a while

Verify before sending

The openers cite specific published work, and the whole approach collapses if a detail is wrong. Your own rules say a first line that misquotes the thing you claim to have read is worse than a generic one.

Check, every time:

  • The affiliation of any academic recipient. arXiv abstract pages often carry no affiliations at all, and the paper's PDF is the place to confirm them. Do not send an academic an email that misplaces them.
  • The link resolves and says what the opener claims it says
  • Any technique id against the source. AML.T0109 is the MCP rug pull; AML.T0051.001 is indirect prompt injection

Sending order, and the reasoning behind it

Openers 2 and 5 first. Both go to people who already published on the topic, and both ask for criticism rather than a meeting. That is a smaller thing to grant and a more useful thing to receive: being told the rug-pull logic or the ATLAS mapping is wrong is worth more right now than a call that goes nowhere. Neither costs anything if ignored.

3 next, with the Splunk SPL attached, because detection engineers respond to artefacts rather than pitches.

1, 4 and 6 last. They are the actual revenue paths and the slowest, and they land better once there is something concrete to reference.

Do not reorder this without saying why. An earlier audit recommended starting with opener 1 and did not engage with the argument above; that was wrong.

After a send

Update the row in the private outreach-log.md with the date and channel. Record the reply verbatim when one arrives, including a rejection. The point of the log is that the next revision is informed by what happened rather than by what sounded good.

What this skill will not do

It will not send anything. No email or messaging tool is available in this session, and outreach on someone's behalf is theirs to press send on regardless. Prepare it, verify it, hand it over.

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.