leiloeiro-mercado

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATIONMETADATA_POISONING
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes real estate market data and user queries which serve as potential ingestion points for untrusted content.
  • Ingestion points: Real estate property descriptions, market data inputs, and ROI calculation parameters provided by the user as described in SKILL.md.
  • Boundary markers: Absent. The skill lacks explicit instructions for the agent to ignore or delimit instructions embedded within processed data.
  • Capability inventory: The skill is knowledge-focused with scripts/governance.py using standard libraries for local logging and rate limiting. No network or high-risk file system operations are present.
  • Sanitization: Absent. There is no evidence of input validation or escaping for the processed data.
  • [DATA_EXFILTRATION]: The SKILL.md file contains a hardcoded absolute local path (C:\Users\renat\skills\agent-orchestrator\scripts\scan_registry.py) in the installation instructions. This leaks the author's local username (renat) and folder structure to the agent context.
  • [METADATA_POISONING]: A discrepancy exists between the platform-provided author identity (sickn33) and the author field declared in the skill metadata (renat), which may lead to confusion regarding the skill's origin.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 10:43 PM
Security Audit — agent-trust-hub — leiloeiro-mercado