daily

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill reads status information from a local daily-state.md file and presents it to the agent, creating a potential vector for indirect prompt injection if the file content is untrusted.
  • Ingestion points: State data is read from daily-state.md via the daily-state.py script and used to generate a welcome-back prompt for the agent in SKILL.md.
  • Boundary markers: The skill lacks explicit delimiters or instructions to prevent the agent from following commands that might be embedded in the status fields of the state file.
  • Capability inventory: The agent can execute Git commands such as stage, commit, and push, and initiate various other workflow-related skills.
  • Sanitization: No input validation or sanitization is performed on the data retrieved from the state file before it is used in the agent's context.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 05:45 AM
Security Audit — agent-trust-hub — daily