tech-contract-negotiation

Pass

Audited by Gen Agent Trust Hub on Oct 29, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external contract drafts and deal parameters provided by the user, which represents a potential ingestion surface for indirect instructions.
  • Ingestion points: The scripts negotiation_position_analyzer.py and deal_complexity_scorer.py ingest content from local text and JSON files provided as command-line arguments.
  • Boundary markers: The scripts do not utilize specific delimiters to isolate potential instructions that might be embedded within a contract draft, relying on standard text processing.
  • Capability inventory: The tools provided are restricted to read-only operations and text analysis via regular expressions. There are no capabilities for file writing, network communication, or subprocess execution of untrusted code found within the scripts.
  • Sanitization: Input content is processed via regex and JSON parsing without specific sanitization for natural language instructions, but the deterministic nature of the scripts limits the impact of such content to the analysis output itself.
  • [COMMAND_EXECUTION]: The skill documentation includes examples of running local Python scripts. These scripts are included in the skill package and perform legitimate, scoped analysis tasks without acquiring elevated privileges or executing shell commands with unsanitized user input.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 29, 2026, 11:32 AM