mlb-signal-emitter
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted signal data (YAML and Markdown) from other skills and persists it to the file system. Ingestion points: Payload received from upstream callers as documented in SKILL.md and resources/methodology.md. Boundary markers: Absent. The skill lacks specific delimiters or ignore instructions for the content of the signals being processed. Capability inventory: The skill uses write_file to write to disk and ensure_dir to create directories. Sanitization: Present. The methodology documentation specifies sanitizing file identifiers and checking that the computed file paths do not escape the designated signals/ directory.
Audit Metadata