hackernews-intel
Pass
Audited by Gen Agent Trust Hub on Apr 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The SKILL.md file contains multiple instructions and examples using hardcoded absolute file paths (e.g., /Users/ksd/Desktop/Varnan_skills/hackernews-intel/). While these paths appear to belong to the author's development environment, their inclusion in the skill's instructions assumes a specific directory structure that may not exist on the user's system, potentially causing command execution failures.- [PROMPT_INJECTION]: The skill processes untrusted data from the Hacker News API and outputs it to the console, creating an indirect prompt injection surface. * Ingestion points: Hacker News stories and comments are fetched via the Algolia API in scripts/monitor-hn.js. * Boundary markers: The terminal output does not use delimiters or provide instructions to the agent to treat the external content as non-instructional data. * Capability inventory: The skill uses Node.js for network and database operations. * Sanitization: Fetched content is not sanitized or escaped before being printed.- [PROMPT_INJECTION]: There is a metadata inconsistency where the skill's author is listed as 'OpenDirectory' in the SKILL.md frontmatter, while the provided author context identifies the vendor as 'Varnan-Tech'.
Audit Metadata