skills/eliasstravik/skills/demo/Gen Agent Trust Hub

demo

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by processing external data into local artifacts.- Ingestion points: The skill resolves a "target" and "demo state root" from provided context (SKILL.md).- Boundary markers: The instructions lack explicit delimiters or warnings to ignore instructions embedded within the target data.- Capability inventory: The skill can write files (index.html), utilize host tools, and manage server processes (SKILL.md).- Sanitization: No sanitization or validation of the input target is required before it is used to stage product behavior.- [COMMAND_EXECUTION]: The agent is directed to use host tools directly and manage long-lived local server processes.- The instructions specify binding the server to 127.0.0.1 on non-common ports to mitigate network exposure.- The workflow involves recording process IDs (PIDs) and restart commands for the local serving environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 03:58 PM
Security Audit — agent-trust-hub — demo