sflo-quick
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user requests to generate executable verification logic, creating a vulnerability surface. \n
- Ingestion points: User-provided authorizing requests processed in
references/gates/frame.mdandreferences/gates/align.md. \n - Boundary markers: Absent. No specific delimiters or safety instructions are defined to separate user intent from verification logic. \n
- Capability inventory: The agent is instructed to execute shell commands, behavioral proofs, and actions in
references/gates/check.md. \n - Sanitization: Absent. No explicit validation or escaping of user-provided criteria is specified before command execution. \n- [COMMAND_EXECUTION]: The workflow requires the agent to execute shell commands and actions to verify criteria as defined in the
CHECKphase (references/gates/check.md). \n- [DYNAMIC_EXECUTION]: The skill generates candidate work products and runs dynamic verification probes at runtime based on the initial task scope defined inreferences/gates/make.mdandreferences/gates/check.md.
Audit Metadata