okr-progress-narrator
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from external files (.md, .txt, .csv) and chat messages to generate narrative summaries, which can lead to the agent following instructions embedded within the data.
- Ingestion points: In
SKILL.md, Step 1 instructions specify reading content from user-provided file paths or chat text. - Boundary markers: Absent. The instructions do not define delimiters or specific warnings to ignore instructions embedded within the processed OKR notes.
- Capability inventory: The skill possesses file-writing capabilities as described in
SKILL.md, Step 6, to createokr-update-YYYY-MM-DD.mdfiles. - Sanitization: Absent. There is no evidence of logic designed to filter or escape potentially malicious prompts contained in the 'notes' or 'context' fields of the input data before they are woven into the output narrative.
Audit Metadata