preparar-sprint0
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests content from external files like TECH-DESIGN.md and ADRs which could theoretically contain malicious instructions designed to influence the output.
- Ingestion points: Reads TECH-DESIGN.md, adrs/ folders, AGENTS.md, and CLAUDE.md.
- Boundary markers: None found in the logic or templates.
- Capability inventory: File reading and writing to SPRINT0.md and AGENTS.md. No command execution or network access.
- Sanitization: No sanitization of ingested text.
- Mitigation: The skill includes a specific workflow step requiring human approval of all findings and documents before completion.
- [SAFE]: No obfuscation, data exfiltration, or persistence mechanisms were detected. The skill's operations are confined to documentation processing within the local workspace, and it explicitly prohibits the implementation of logic or network communication.
Audit Metadata