spec-flow-analyzer
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, unauthorized system commands, or network operations were detected in the skill instructions.
- [PROMPT_INJECTION]: The skill operates as an ingestion surface for indirect prompt injection because it is designed to analyze untrusted data provided by users (such as specifications and feature plans). The risk is mitigated by the fact that the agent lacks access to external tools, network capabilities, or local file systems.
- Ingestion points: User-provided specification and planning text (SKILL.md).
- Boundary markers: Absent; user content is not explicitly delimited from system instructions.
- Capability inventory: No tools, code execution methods, or network access are available to this skill.
- Sanitization: No sanitization or input validation is performed on the provided documentation.
Audit Metadata