skills/octavehq/lfgtm/insights/Gen Agent Trust Hub

insights

Pass

Audited by Gen Agent Trust Hub on Feb 27, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through its data processing flow.
  • Ingestion points: Untrusted data enters the agent context via call transcripts and email contents retrieved through list_events and get_event_detail as specified in Step 2 and Step 4 of SKILL.md.
  • Boundary markers: The instructions do not define delimiters or provide specific warnings to the agent to ignore instructions embedded within the conversation transcripts.
  • Capability inventory: The skill is granted the capability to modify organizational knowledge via the update_entity tool, which can overwrite personas and playbooks based on processed insights (Step 5).
  • Sanitization: There is no evidence of content sanitization, escaping, or validation of the retrieved transcript data before it is interpolated into the reasoning process for library updates.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 27, 2026, 03:09 AM