rss-digest
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and installs a command-line utility named
feedfrom the author's GitHub repository (github.com/odysseus0/feed) using standard package managers (Homebrew and Go). It also downloads a starter list of feeds (OPML file) from the same repository to populate the tool if no feeds are present. - [COMMAND_EXECUTION]: The skill performs local operations using the
feedCLI, such as fetching entries, searching the local database, and updating the read status of posts. - [INDIRECT_PROMPT_INJECTION]: The skill functions by ingesting untrusted data from external RSS feeds, which presents a surface for indirect prompt injection.
- Ingestion points: The agent fetches titles, summaries, and full article content from remote RSS feeds (SKILL.md).
- Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded commands within the fetched RSS content.
- Capability inventory: The skill uses the
feedCLI for data management and implies the use of separate web-fetching tools for retrieving full article text. - Sanitization: No specific sanitization or filtering logic is provided in the instructions for the external feed content.
Audit Metadata