swot-from-signals

Pass

Audited by Gen Agent Trust Hub on Jul 14, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data sources including social media mentions, web search trends, and third-party reviews.
  • Ingestion points: Data is loaded from pricing_scrape.csv, serp_trends.json, social_mentions.json, and dataset_imports/*.csv (Naver, Coupang, Trustpilot).
  • Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded prompts in the source data.
  • Capability inventory: The skill uses pandas for local file processing and writes output to a local JSON file.
  • Sanitization: The instructions include steps for schema, missing value, and duplicate checks, which provide data integrity but not prompt sanitization. However, the structured output format and objective constraints (no speculative language) provide some mitigation against unintentional obedience to data-embedded instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 14, 2026, 04:31 PM
Security Audit — agent-trust-hub — swot-from-signals