web-content-fetcher

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from arbitrary external URLs, which acts as a vector for indirect prompt injection.
  • Ingestion points: The skill fetches content from the web via the Jina Reader API and the scrapling library within scripts/fetch.py.
  • Boundary markers: There are no explicit delimiters or protective instructions used to isolate the fetched external content from the agent's primary instructions.
  • Capability inventory: The skill maintains capabilities for network access and local script execution.
  • Sanitization: While the skill converts HTML to Markdown, this process does not filter out natural language instructions that could influence the agent's behavior.
  • [COMMAND_EXECUTION]: The skill utilizes a shell execution pattern in SKILL.md (exec: python3 scripts/fetch.py <url>) that directly incorporates a user-supplied URL as a command-line argument. This creates a surface for command injection if the execution environment does not adequately sanitize shell metacharacters.
  • [EXTERNAL_DOWNLOADS]: The skill's operational setup involves the scrapling install command, which downloads browser binaries and related dependencies from external remote servers.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 05:23 PM
Security Audit — agent-trust-hub — web-content-fetcher