gtm-operator-skills
AGENTS.md3 starsChanged 24 days agoMITView on GitHub
Instructions for any coding agent (Claude Code, Codex, Cursor, Grok, and the rest) working with this repo or the skills in it. Nineteen GTM skill packs under skills/<pack>/, each with one skill, a Python scorer, evals and a worked sample. Plus suite/ (the gtm plugin: /gtm:setup, /gtm:next), company-context-pack/, pick-gtm-tools/, skill-audit/ and build-pack/. Claude Code only: Do not commit or distribute CSV files, proprietary datasets or bulk exports in any format. Never repackage withdrawn rows as JSON, images, Markdown or XLSX.…
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 32,671 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository7 other files it gives its agents.
- skills/cold-email/AGENTS.mdAGENTS.md
- skills/evp/AGENTS.mdAGENTS.md
- skills/founder-brand/AGENTS.mdAGENTS.md
- skills/pricing/AGENTS.mdAGENTS.md
- skills/psp/AGENTS.mdAGENTS.md
- build-packSkill
- company-context-packSkill
The file
We split this AGENTS.md into the rules it gives an agent
4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 24
Run the pack's scorer on every draft and show the user each flag.
- Distinctive: line 27
Never invent a signal, a result, a quote or a customer name.
- Very specific: line 34
Do not edit a pack's framework text without running its evals: python3 -m unittest discover -s tests inside the pack folder.
- Very specific: line 35
Before a change to a manifest: claude plugin validate --strict .
≈ “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.
Files
AGENTS.md
Instructions for any coding agent (Claude Code, Codex, Cursor, Grok, and the rest) working with this repo or the skills in it.
What is here
Nineteen GTM skill packs under skills/<pack>/, each with one skill, a Python scorer, evals and a worked sample. Plus suite/ (the gtm plugin: /gtm:setup, /gtm:next), company-context-pack/, pick-gtm-tools/, skill-audit/ and build-pack/.
Install
npx skills add cmj-hub/gtm-operator-skills --all -g --full-depthClaude Code only:
/plugin marketplace add cmj-hub/gtm-operator-skills
/plugin install gtm@gtm-operator-skillsRules for using a pack
- Run the pack's scorer on every draft and show the user each flag. Exit 0 passes, 1 refuses, 2 is bad input. The last line of the output is
Next:. - Drafts go in
gtm/at the project root, one file per pack. Founder posts go indrafts/. Write nowhere else. brand-config.jsonandSOUL.mdare shared by every pack. Merge into them. Never overwrite a filled value without asking.- Never invent a signal, a result, a quote or a customer name. A claim needs a source and a date, or it stays out.
- Nothing in a pack sends, posts or publishes. A human does that.
- The scorers are standard-library Python. They need no key and make no network call.
Rules for editing this repo
- A skill description is trigger phrases a user would type, then
For X, see Y.boundaries. Keep each pack under 400 always-on tokens. - Do not edit a pack's framework text without running its evals:
python3 -m unittest discover -s testsinside the pack folder. - Before a change to a manifest:
claude plugin validate --strict .from the repo root. - Scope
WriteandEditinallowed-toolsto the folder the skill writes, for exampleWrite(./gtm/**).
Public sample and data policy (Jay, 2026-10-23)
Do not commit or distribute CSV files, proprietary datasets or bulk exports in any format. Never repackage withdrawn rows as JSON, images, Markdown or XLSX. A CSV for Jay requires his explicit approval after inspecting its exact contents. Preserve operator-provided private-input tooling. Tests must use native fictional dictionaries and mocked I/O without creating CSV files or serializing CSV samples.
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.

