escribir-contrato
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill's primary function is text analysis and document generation. It does not include executable scripts, network requests, or privileged system commands.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external user stories and design documents, which is a common attack surface for indirect prompt injection.
- Ingestion points: Reads User Stories (HU) via MCP or local files, and processes the
TECH-DESIGN.mdfile. - Boundary markers: The workflow includes strict requirements to cite the origin of each line and explicitly labels non-sourced information as 'DECISIÓN PROPUESTA'.
- Capability inventory: Capabilities are restricted to reading project documentation and writing markdown files to the
contratos/directory. No subprocess or network tools are utilized. - Sanitization: The skill enforces a structured template and mandates that the agent stops for human approval of the generated contract and any open questions.
Audit Metadata