mastodon-mcp-cli
AGENTS.md0 starsChanged 39 days agoMITView on GitHub
For agents editing this repository. Users read the README. Driving the server is SKILL.md. Commit as n@navid.me. Never pass -c user.email=. The global config is correct and the override is the bug: the wrong address credits a blocked account and the Contributors panel reads 0. Writes are on by default. MASTODON_READ_ONLY=1 is the opt-out and it works by not registering the write tools, not by refusing at call time. Approval on operations that reach other people only. Posting, threads, editing,…
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 31,774 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.
The file
We split this AGENTS.md 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 19
Commit as
n@navid.me: Never pass -c user.email=. - Distinctive: line 29
Not favourites, boosts, follows or mutes: each is one click to undo, and confirming everything trains the reflex that makes the …
≈ “Common” is an estimate: at least 1 in 5 of the 40,246 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
Working on mastodon-mcp-cli
For agents editing this repository. Users read the README. Driving the server is
SKILL.md.
Layout
src/app.ts the Slipway app: tools, settings, doctor, login, logout. Slipway owns both surfaces, the guard and the audit log
src/guide.ts server instructions, resources and prompts
src/api/ client, errors
src/auth/ app registration and sign-in, per instance
src/content/ status text, media, content warnings
src/tools/ one module per group, registered in tools/index.tsNon-negotiables
Commit as n@navid.me. Never pass -c user.email=. The global config is
correct and the override is the bug: the wrong address credits a blocked account
and the Contributors panel reads 0.
Writes are on by default. MASTODON_READ_ONLY=1 is the opt-out and it works
by not registering the write tools, not by refusing at call time.
Approval on operations that reach other people only. Posting, threads,
editing, deleting, blocking, reporting, poll votes. Over MCP a person approves
each in the client; confirm: true counts only where the client cannot ask.
Not favourites, boosts, follows or mutes: each is one click to undo, and confirming everything trains the reflex
that makes the confirmation on a delete worthless.
There is no developer portal. Mastodon has no central place to register an
app, so the server registers its own application against whichever instance the
user names, then signs in. That is why setup is one command and why the code
carries an auth/ directory the other servers do not need.
Instance differences are the platform. Character limits, poll limits, media
counts and available features vary per instance and are read from
/api/v1/instance rather than assumed. Never hardcode 500 characters.
Every anticipated failure is a MastodonError from api/errors.ts, with a
message the model can act on. Slipway maps its HTTP status to the exit code and
keeps the message. A plain Error keeps its message too, but exits 1,
unexpected, unless its words match a known failure.
Before claiming it works
npm run build && npm test && npm run typecheck
npx @modelcontextprotocol/inspector node dist/index.jsDoes 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.

