everyinfra-agent-plugins
CLAUDE.md347 starsChanged 2 months agoApache-2.0View on GitHub
This is the public integration package for EveryInfra. It is intentionally independent from the private gateway repository. - Keep MCP, Plugin, registered App/Connector, and Marketplace status distinct. - Treat live MCP discovery and public catalogs as runtime truth. - Never add credentials, customer data, private fields, or internal implementation details. - Do not describe EveryNumber, EveryMail, or EveryProxy as MCP-enabled until production discovery and parity tests prove it. - Paid calls and external side effects must remain explicit and bounded.
In this repositoryOne other file it gives its agents.
- llms.txtllms.txt
The file
We split this CLAUDE.md into the rules it gives an agent
8 rules. About 0 are common, 6 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈6
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- could be automatic: a setting could enforce it
- Distinctive: line 8
Keep MCP, Plugin, registered App/Connector, and Marketplace status distinct.
- Distinctive: line 9
Treat live MCP discovery and public catalogs as runtime truth.
- Distinctive: line 10
Never add credentials, customer data, private fields, or internal implementation details.
- Distinctive: line 11
Do not describe EveryNumber, EveryMail, or EveryProxy as MCP-enabled until production discovery and parity tests prove it.
- Distinctive: line 24
Keep version-bearing manifests and marketplace entries aligned for releases.
- Very specific: line 25
Update README.md, llms.txt, and docs/ai-discovery.md when the public capability boundary changes.
- Very specific: line 27
Run python3 scripts/validate.py and the available host validators before handoff.
- Distinctive: line 28
Do not commit, push, publish, or submit to a marketplace without explicit user authorization.
≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.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
EveryInfra agent plugin repository
This is the public integration package for EveryInfra. It is intentionally independent from the private gateway repository.
Stable boundaries
- Keep MCP, Plugin, registered App/Connector, and Marketplace status distinct.
- Treat live MCP discovery and public catalogs as runtime truth.
- Never add credentials, customer data, private fields, or internal implementation details.
- Do not describe EveryNumber, EveryMail, or EveryProxy as MCP-enabled until production discovery and parity tests prove it.
- Paid calls and external side effects must remain explicit and bounded.
Change and validation workflow
- Shared skills live only under
plugins/everyinfra/skills/; host manifests are thin adapters. workbuddy/holds WorkBuddy connectors with Chinese host-adapted skills. When a shared skill's rule changes, update the matching WorkBuddy skill too; generated files come fromscripts/build_workbuddy.py render, andscripts/build_workbuddy.py checkmust pass.host-packages/holds manifests and listing metadata for markets that take uploads (MiniMax, Qoder/QoderWork, ClawHub);scripts/build_host_packages.py checkmust pass, andbuildwrites the packages todist/without submitting them.- Keep version-bearing manifests and marketplace entries aligned for releases.
- Update
README.md,llms.txt, anddocs/ai-discovery.mdwhen the public capability boundary changes. - Run
python3 scripts/validate.pyand the available host validators before handoff. - Do not commit, push, publish, or submit to a marketplace without explicit user authorization.
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.

