agentleFS
selamy-labsselamy-labs/agent-skills/skills/connector-readiness/SKILL.md

connector-readiness

Skill0 starsChanged 6 days agoMITView on GitHub

Use when installing, repairing, or verifying an MCP or other authenticated connector, especially when configuration, authentication, quota, and real artifact access can be mistaken for one another.

The file

We split this skill into the rules it gives an agent

10 rules. About 0 are common, 10 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈10

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Treat connector readiness as three separate gates.

    line 8
  2. Distinctive:

    Run the connector's identity or health command as the same operating-system user and client that owns the work.

    line 19
  3. Distinctive:

    Run the smallest representative read-only operation against the exact artifact, object, node, or repository needed by the task.

    line 23
  4. Distinctive:

    Quota or plan limit: stop retrying. Preserve the exact server response, check current official plan guidance, and distinguish a server-side …

    line 37
  5. Distinctive:

    Client or model incompatibility: update the client, start a fresh process, and rerun the artifact-capability gate.

    line 40
  6. Distinctive:

    Use the provider's design-to-code skill before its primary context tool.

    line 49
  7. Distinctive:

    Prefer the primary design-context call.

    line 51
  8. Distinctive:

    If a live call is quota-blocked, label any accepted repository snapshot or export as a fallback with immutable lineage.

    line 53
  9. Distinctive:

    Verify both the client's connector list and a representative tool call.

    line 62
  10. Distinctive:

    Call the connector ready only when the intended worker returns a real payload from the required artifact.

    line 70

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Connector Readiness

Treat connector readiness as three separate gates. Passing an earlier gate does not imply that a later gate works.

Readiness Gates

  1. Configured
    • Confirm the plugin or server is enabled in the client that will do the work.
    • Confirm the endpoint, transport, and client version.
    • Restart or start a fresh client process after changing plugin or MCP configuration when the client loads tools only at startup.
  2. Authenticated
    • Run the connector's identity or health command as the same operating-system user and client that owns the work.
    • Record the account and plan or seat without exposing tokens.
  3. Artifact-capable
    • Run the smallest representative read-only operation against the exact artifact, object, node, or repository needed by the task.
    • Require a real payload such as a name, dimensions, revision, or metadata. A connected status, successful OAuth callback, or tool listing is insufficient.

Error Classification

  • Authentication or permission: re-authenticate, confirm the account, and verify that account can access the target artifact.
  • Invalid argument: re-extract exact identifiers from the canonical link and try a documented orientation call. If unrelated exact artifacts fail too, investigate the client wrapper or server schema instead of repeatedly guessing identifiers.
  • Quota or plan limit: stop retrying. Preserve the exact server response, check current official plan guidance, and distinguish a server-side quota from broken local configuration.
  • Client or model incompatibility: update the client, start a fresh process, and rerun the artifact-capability gate.
  • Timeout or oversized response: narrow the operation to a smaller object without substituting an unauthoritative source.

Design-Source Connectors

For design-to-code work:

  • Use the provider's design-to-code skill before its primary context tool.
  • Copy an exact frame or layer link; do not guess a node identifier.
  • Prefer the primary design-context call. Use identity, metadata, or screenshot calls only for diagnosis and orientation.
  • If a live call is quota-blocked, label any accepted repository snapshot or export as a fallback with immutable lineage. Do not describe fallback data as a successful live connector read.

Remote Worker Verification

Configuration may differ by user, client, host, or already-running session.

  • Inspect readiness under the worker's real user and login environment.
  • Verify both the client's connector list and a representative tool call.
  • Inspect the active worker transcript or trace for the exact call and result. A healthy connector in a different client does not prove the worker used it.
  • Record the client, version, user, session, target identifier, operation, and observed outcome.

Done

Call the connector ready only when the intended worker returns a real payload from the required artifact. If quota, permissions, or external state prevents that, report the connector as configured or authenticated but not artifact-capable.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.