competitor-signals

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill connects to well-known public APIs, including Product Hunt (GraphQL) and Hacker News (Algolia), to retrieve engagement data such as comments and upvotes.
  • [COMMAND_EXECUTION]: Executes a local Python utility (scripts/competitor_signals.py) to process retrieved data, calculate signal scores, and generate CSV output.
  • [PROMPT_INJECTION]: The skill processes untrusted user-generated content from external platforms (Product Hunt and Hacker News comments). This represents a surface for indirect prompt injection where malicious content in comments could attempt to influence the agent's analysis.
  • Ingestion points: Public comments and website text retrieved via the WebFetch tool and the Python processing script.
  • Boundary markers: None explicitly defined in the prompt instructions when the agent evaluates the processed results.
  • Capability inventory: The agent utilizes Bash, Write, and Edit tools to manage the configuration and output files.
  • Sanitization: The Python script implements basic HTML tag stripping to clean scraped content before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 03:06 PM
Security Audit — agent-trust-hub — competitor-signals