The Agent Skills Directory

[SAFE]: The skill enforces security best practices by explicitly instructing the agent never to log or print access tokens, client secrets, refresh tokens, or board content. It also mandates the use of OAuth 2.0 with the narrowest possible scopes.
[SAFE]: External references are limited to official Miro developer documentation and status pages, which are well-known and trusted services.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it reads repository files using Read, Glob, and Grep and has the capability to Write or Edit files.
Ingestion points: Files within the analyzed repository read via Read, Glob, and Grep (referenced in SKILL.md).
Boundary markers: Absent; the instructions do not specify using delimiters or "ignore" instructions for analyzed code.
Capability inventory: Write and Edit for modifying code, WebFetch for external documentation access.
Sanitization: The skill emphasizes redacting output telemetry (e.g., tenant hashes, credential fields) but does not explicitly specify sanitization of input data before processing.

miro-observability