ai-daily-digest
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches content from 90+ hardcoded RSS feeds belonging to well-known technology blogs and individuals (e.g., Simon Willison, Jeff Geerling, Andrej Karpathy). These are trusted sources for the skill's primary purpose.
- [COMMAND_EXECUTION]: The skill executes a local TypeScript script using
bun(vianpx -y bun) to perform the RSS fetching. It also utilizes a local installation of Google Chrome in headless mode to convert the Markdown report to a PDF. These operations are restricted to the local environment and are standard for report generation tasks. - [INDIRECT_PROMPT_INJECTION]: As the skill ingests external content from RSS feeds to provide summaries, it possesses an indirect prompt injection surface. However, the risk is mitigated by:
- Ingestion points: RSS feed titles and descriptions in
scripts/digest.ts. - Boundary markers: The prompt templates in
scripts/digest.ts(buildScoringPromptandbuildSummaryPrompt) use clear delimiters and structured indexing to separate different articles. - Capability inventory: While the skill can write files and execute local commands (Chrome), these are used for output generation rather than executing content derived from the feeds.
- Sanitization: The script includes a
stripHtmlfunction to sanitize the text before it is processed by the AI.
Audit Metadata