keel-ops
Skill0 starsChanged 27 days agoMITView on GitHub
Operations & connectors lane router. Routes memory, workspace connectors, and meta/skills-management to the right tool — claude-mem, granola, slack, notion, zapier MCPs, plus find-skills and the claude-skills library. Defers to keel-v2. Use for persistent memory, pulling meeting notes, reading/writing Slack or Notion, wide app automation, or discovering/installing new skills.
In this repository16 other files it gives its agents.
- skills/keel-v2/CLAUDE.mdCLAUDE.md
- doc-syncSkill
- fused-landingSkill
- keel-contentSkill
- keel-designSkill
- keel-devSkill
- keel-financeSkill
- keel-legalSkill
- keel-marketingSkill
- keel-qaSkill
- keel-researchSkill
- keel-secureSkill
- keel-setupSkill
- keel-v2Skill
- keel-verifySkill
- lazy-loopSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is 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.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 18
Run the keel-v2 scan. Connector *reads* are usually FAST/no-artifact.
- Distinctive: line 38
Don't duplicate it in this lane.
≈ “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.
Files
Keel Ops — operations & connectors lane router
Index, not an implementation. Routes to a connector or meta tool; most
here are MCP servers (wired in .mcp.json), not skills.
Deference — keel-v2 is the arbiter
Run the keel-v2 scan. Connector reads are usually FAST/no-artifact. Connector writes are irreversible/outward-facing (surface 6) → HEAVY, and the harness send/publish/settings gates apply: posting to Slack, writing a Notion page, or firing a Zapier action is confirmed with the human first, never auto-run.
Dispatch table
| When the task is… | Dispatch to | Type | Source |
| Persistent cross-session memory (capture, compress, re-inject context) | claude-mem | skill/tool | thedotmack/claude-mem |
| Pull meeting transcripts / notes into context | granola | MCP | granola.ai |
| Read/search or post Slack messages, threads, canvases | slack | MCP | slackapi/slack-mcp-plugin |
| Read/write Notion pages and databases | notion | MCP | makenotion/notion-mcp-server |
| Fire an action across 9,000+ apps in plain English | zapier | MCP | zapier/zapier-mcp |
| Discover & install new skills from the open ecosystem | find-skills | skill | vercel-labs/skills |
| Reach into a large multi-platform skills library | claude-skills | library | alirezarezvani/claude-skills |
Token thrift note
Output-token compression is not here — it's your existing lazy-loop
skill (/lazy), which already wraps caveman + ponytail and defers to keel-v2.
Don't duplicate it in this lane.
Write-action gate (non-negotiable)
Any Slack post, Notion write, Zapier trigger, or other outward/irreversible connector action: state exactly what will be sent/changed and to where, then wait for an explicit human yes. One approval does not authorize the next.
If the target isn't installed
Say so; MCP connectors need credentials in .mcp.json (see .mcp.json.template
install.sh mcp) and cannot be faked. Install a skill lane:./install.sh ops.
Sources: thedotmack/claude-mem · granola.ai · slackapi/slack-mcp-plugin · makenotion/notion-mcp-server · zapier/zapier-mcp · vercel-labs/skills · alirezarezvani/claude-skills. Upstreams unmodified.
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.