defuddle
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructions include installing the
defuddlepackage from the NPM registry vianpm install -g defuddle. This package is authored by the skill developer. - [COMMAND_EXECUTION]: The skill utilizes the
defuddlecommand-line interface to parse external URLs and includes options for writing output to the local file system (e.g.,defuddle parse <url> --md -o content.md). - [INDIRECT_PROMPT_INJECTION]: As the skill is designed to fetch and process content from arbitrary web pages, it possesses an inherent surface for indirect prompt injection where instructions on a parsed page could attempt to influence the agent's behavior.
- Ingestion points:
SKILL.md(via thedefuddle parse <url>command). - Boundary markers: None specified in the instructions.
- Capability inventory: Command execution and file system write capabilities.
- Sanitization: No sanitization or filtering of the parsed content is described in the skill instructions.
Audit Metadata