cmd-rss-feed-generator
Warn
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: MEDIUMDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill workflow requires the agent to generate Python source code (scripts in
feed_generators/) based on the analyzed structure of external sites and then execute that code using theuv runcommand. This pattern involves executing code influenced by untrusted external data within the system environment. - [INDIRECT_PROMPT_INJECTION]: The agent is tasked with fetching and analyzing HTML content from external blog URLs to identify data selectors. Malicious instructions embedded in the source HTML could attempt to subvert the agent's logic during script creation or metadata updates. 1. Ingestion points: External HTML content retrieved through
requestsorselenium. 2. Boundary markers: None specified for the untrusted data. 3. Capability inventory: File system write access and subprocess execution viauv runandmake. 4. Sanitization: No sanitization or content validation for the fetched HTML is described. - [COMMAND_EXECUTION]: The skill directs the agent to use various shell commands and tools, including
uv run,make,cat, andtail, to manage, test, and validate the feed generation process.
Audit Metadata