competitive-intelligence-gathering

Pass

Audited by Gen Agent Trust Hub on Mar 18, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill demonstrates an unsafe pattern for processing external data using an LLM. An attacker (prospect) could inject instructions into a conversation that override the extraction rules.\n
  • Ingestion points: The extract_intel_with_llm function in SKILL.md accepts untrusted conversation text.\n
  • Boundary markers: The prompt template in SKILL.md lacks clear delimiters (e.g., XML tags or triple backticks) to separate instructions from the {format_conversation(conversation)} data.\n
  • Capability inventory: The skill possesses the capability to write extracted data to a persistent database (CompetitiveIntelDB), trigger alerts to internal teams (trigger_alert_if_needed), and modify internal battlecards (update_battlecard_from_intel).\n
  • Sanitization: No sanitization, escaping, or filtering of the input text is present in the illustrated code.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 18, 2026, 10:46 PM