knowledge-synthesis
Pass
Audited by Gen Agent Trust Hub on Apr 6, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill instructions involve processing and summarizing untrusted data from external sources (e.g., chat logs, emails, cloud documents). This creates a vulnerability surface for indirect prompt injection, as malicious content within those sources could influence the agent's behavior.\n
- Ingestion points: The skill explicitly processes results from chat, email, cloud storage, and project trackers (SKILL.md).\n
- Boundary markers: There are no instructions for the agent to use delimiters or ignore potentially malicious embedded commands in the source data.\n
- Capability inventory: The skill environment allows access to
run_shell_commandandread_file(SKILL.md).\n - Sanitization: The instructions do not include steps for sanitizing, validating, or escaping external content before synthesis.\n- [NO_CODE]: The skill contains only instructional markdown and guidelines; it does not include any scripts, automated commands, or binary executables.
Audit Metadata