The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it is designed to audit source code diffs and files which may contain untrusted data.
Ingestion points: Audits code diffs and file contents provided to the audit mode in SKILL.md and rules/audit-checklist.md.
Boundary markers: Absent; the skill does not define specific delimiters or safety instructions to prevent the agent from obeying instructions embedded in code comments.
Capability inventory: The skill can update persistent repository configuration through persistent-memory and trigger shell command execution via the observe-run skill.
Sanitization: Absent; there is no logic described to sanitize or validate strings or comments extracted from audited files before they are processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill references and provides instructions for using external tools and resources from well-known technology organizations.
References the weaver CLI and semantic convention registries from the OpenTelemetry project's official GitHub repositories.
Suggests the use of specialized instrumentation skills provided by the dash0hq GitHub organization.
These references target established services and do not involve untrusted or malicious sources.

measurable