fde-desk
AGENTS.md0 starsChanged 48 days agoMITView on GitHub
Registry of desk agents. Each agent lives in .claude/agents/<name>.md and is invoked by name or through a command.
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 repository12 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- clay-table-designSkill
- competitor-teardownSkill
- deploy-surface-checkSkill
- fde-handoff-packSkill
- hermes-herdr-loopSkill
- icp-scorecardSkill
- jev-typesafe-briefSkill
- n8n-flow-scaffoldSkill
- signal-to-sequenceSkill
- trigger-job-patternSkill
- weekly-pipeline-pulseSkill
The file
We split this AGENTS.md into the rules it gives an agent
3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
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 21
Always run ship-reviewer before /desk-close when artifacts will be handed to a human or CRM.
- Very specific: line 23
Run infra-ops (and deploy-surface-check) before blaming Clay or n8n when a hosted surface is red.
- Very specific: line 28
Add .claude/agents/<kebab-name>.md with role, inputs, outputs, and stop conditions.
≈ “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
Registry of desk agents. Each agent lives in .claude/agents/<name>.md and is invoked by name or through a command.
Agents
| Agent | Role | Typical inputs | Typical outputs |
signal-scout | Find and normalize buying triggers | ICP notes, watchlist domains, Clay signal feeds | Signal board markdown, ranked trigger list |
clay-enricher | Design and run Clay enrichment batches | Table schema, row sample, enrichment goals | Clean CSV/table plan, field map, QA notes |
account-mapper | Map buying committee and whitespace | Account list, org chart hints, CRM notes | Account map, persona coverage gaps |
outreach-drafter | Draft outbound that cites signals | Signal + account map + offer notes | Reviewable email/LinkedIn drafts |
call-prep | Prep discovery and demo calls | Meeting context, account map, open questions | Call brief, objection list, success criteria |
ship-reviewer | Gate quality before anything leaves the desk | Drafts, packs, enrichment diffs | Pass/fail review with fix list |
infra-ops | Triage Render / Railway / optional Oracle A1 | Symptoms, deploy notes, stack-status | Triage note, status colors, fix order |
Coordination
- Scout and enrich before draft.
- Map accounts before call prep.
- Always run
ship-reviewerbefore/desk-closewhen artifacts will be handed to a human or CRM. - Agents draft; humans approve; hooks enforce the send fence.
- Run
infra-ops(anddeploy-surface-check) before blaming Clay or n8n when a hosted surface is red. - Overnight or multi-hour work goes through Hermes/Herdr with a Jev brief, not an open interactive chat left unattended.
Adding an agent
- Add
.claude/agents/<kebab-name>.mdwith role, inputs, outputs, and stop conditions. - Register it in this file.
- Wire it into a command or skill if it is part of the default loop.
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.

