agentleFS
Sign inStart free
calebnewtonusccalebnewtonusc/Chewbacca/skills/call-coach/SKILL.md

call-coach

Skill8 starsChanged 3 months agoMITView on GitHub

Debrief a sales or client call from its transcript or recording. Use when the user shares a call transcript, mentions a call that went well or badly, asks how they did, wants to prep for a follow-up, or asks what they missed. Also use on webinar, pitch and demo transcripts, and before a meeting to build a battle card on who is on the other side.

1 other project uses a version of this.

Adopted by

Most recently taken up by

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. Very specific:

    Two hard lines: Never invent a case study.

    line 51
  2. Distinctive:

    Not a verdict on the person.

    line 57
  3. Distinctive:

    When a debrief is right, keep the call.

    line 63

≈ “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

call-coach

Reads a call and returns a debrief while the call is still fresh. The point is the loop: feedback within minutes changes the next call, feedback in a week changes nothing.

Getting the transcript

Granola, Zoom, Fathom and Otter all export text. A recording on disk goes through whisper <file> --model small.en --language en --output_format txt. A Mac call with no recording is a dead end, so say that rather than guessing at what was said.

The debrief

Return these, in this order, and nothing else unless asked:

  1. What happened. Five sentences, no more. What they want, what they objected to, what was agreed.
  2. Talk ratio. Who spoke more, roughly, as a percentage. The buyer should be talking more than the seller. If the seller is over half, that is the headline finding and it goes first.
  3. Buying signals, quoted. A signal is something they said, not a vibe.
  4. Key moments, with the line that made each one matter.
  5. Missed opportunities. The question that should have been asked, at the point it should have been asked.
  6. Score out of 10, with the two things that cost the most points.
  7. Follow-up email, in the user's voice, ready to send. Draft only.

Scoring lenses

Score against negotiation and offer frameworks: calibrated questions, labelling, tactical empathy, and whether the offer's value was made concrete before price came up. Name which lens each deduction comes from so the user can argue with it.

Swap the lenses for whichever practitioners the user actually trusts. Ask once, then remember the answer in the second brain rather than asking again.

Before the call, not after

Same skill, different direction. Given a name and a company, build a battle card: who this person is, what they have built, how to open, which proof to bring, and the five objections they will raise with a response to each.

Two hard lines. Never invent a case study. Never invent a client name. An unverifiable specific goes in as [NEED: ...], never as a plausible value, because the one person who notices is the one who was there.

What this is not

Not a verdict on the person. A bad call is usually a structural problem: no discovery, price before value, or a question that never got asked. Name the structure, not the character.

Feeding it back

When a debrief is right, keep the call. A corpus of the user's own calls that closed beats any generic framework, and this skill gets better by reading them rather than by being rewritten.

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.

Check your setup, free