The Agent Skills Directory

[NO_CODE]: The skill is composed entirely of markdown instructions and YAML metadata. It does not include any executable scripts, binaries, or code files.
[SAFE]: The skill restricts the agent's capabilities to Read, Glob, and Grep tools. This read-only access is appropriate for evidence validation and prevents the skill from modifying the system or exfiltrating data.
[PROMPT_INJECTION]: The skill's primary function involves analyzing untrusted external artifacts, which creates a surface for indirect prompt injection. A malicious evidence file could contain instructions designed to deceive the auditor.
Ingestion points: Audit evidence files (screenshots, logs, configs, policies, spreadsheets) accessed via Read, Glob, and Grep as defined in SKILL.md.
Boundary markers: Absent; no instructions are provided in SKILL.md to the agent to treat ingested data as untrusted or to utilize specific delimiters.
Capability inventory: The skill is limited to file reading and searching using the Read, Glob, and Grep tools listed in SKILL.md.
Sanitization: Absent; the skill does not specify any validation or sanitization steps for the data contained within the artifacts in SKILL.md.

evidence-validator