ucs-page

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external demo HTML files and API specifications which could contain malicious instructions designed to manipulate the agent's behavior. \n
  • Ingestion points: The agent reads files from docs/product/demo/, docs/specs/API/, and docs/specs/grpc/.\n
  • Boundary markers: The instructions do not specify any delimiters or safety markers to help the subagents distinguish between data and potential instructions within the ingested content.\n
  • Capability inventory: Subagents are granted the capability to write new files to the docs/specs/page-UCS/ directory.\n
  • Sanitization: There is no evidence of validation or sanitization being applied to the ingested content before it is passed to the LLM context.\n- [COMMAND_EXECUTION]: The skill executes a local Python script scripts/slice-components.py via a shell command string that includes filenames from the docs/product/demo/ directory as arguments, which could present a risk of command injection if filenames are maliciously crafted.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 02:49 PM
Security Audit — agent-trust-hub — ucs-page