competitive-intel

Pass

Audited by Gen Agent Trust Hub on Oct 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes competitor data from external JSON files to generate strategic artifacts like battlecards and board summaries.\n
  • Ingestion points: SKILL.md defines several JSON files (competitors.json, tracking_data.json, swot_data.json) as the primary input sources for analysis.\n
  • Boundary markers: The instructions do not specify any delimiters or warnings to prevent the model from obeying instructions that might be embedded within the competitive data entries.\n
  • Capability inventory: The skill leverages included Python scripts (scripts/competitor_tracker.py, scripts/market_landscape_mapper.py, scripts/swot_analyzer.py) to read and transform local file data into readable reports.\n
  • Sanitization: There is no mechanism described or implemented to sanitize strings from the input JSON files before they are interpolated into the prompt templates used for generating sales battlecards or executive summaries.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 27, 2026, 11:10 PM