activity-summary

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill instructs the agent to read and process external activity digests located in the $MOUNT/digests/ directory. These files are generated from external sources (GitHub, Linear, Slack, etc.), which creates a surface for indirect prompt injection if those sources contain malicious instructions designed to influence the agent's behavior.\n
  • Ingestion points: The agent is directed to read files such as $MOUNT/digests/yesterday.md, $MOUNT/digests/today.md, and date-specific Markdown files.\n
  • Boundary markers: There are no explicit instructions or delimiters defined to isolate the content of these digests or to warn the agent against executing embedded commands.\n
  • Capability inventory: The agent uses standard file system tools like ls, cat, and head to retrieve this data.\n
  • Sanitization: No sanitization or validation of the digest content is performed before the agent processes and summarizes it.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 10:33 PM
Security Audit — agent-trust-hub — activity-summary