skills/howells/skills/deslop/Gen Agent Trust Hub

deslop

Pass

Audited by Gen Agent Trust Hub on Sep 5, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted prose from user-provided file paths and URLs without explicit sanitization or boundary markers, creating a potential attack surface for malicious instructions embedded in the data to influence the agent's behavior.
  • Ingestion points: The 'Input Handling' section in SKILL.md specifies that the agent should read local files and fetch page content from URLs provided in the input.
  • Boundary markers: The instructions do not define specific delimiters or 'ignore instructions' wrappers for the ingested content, which could lead the agent to interpret data as commands.
  • Capability inventory: The skill utilizes file system read access and network browsing tools to retrieve text for analysis.
  • Sanitization: There are no explicit instructions for filtering or validating the retrieved content before it is processed by the LLM.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 5, 2026, 07:03 AM
Security Audit — agent-trust-hub — deslop