skills/ilyautov/sbis-mcp-ru/sbis-mcp/Gen Agent Trust Hub

sbis-mcp

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of the sbis-mcp-ru package from a public registry via the uvx command.
  • [REMOTE_CODE_EXECUTION]: By installing and running the sbis-mcp server, the AI assistant executes external code that facilitates communication with the SBIS API. This server operates with the provided SBIS_SESSION_ID and has the capability to perform actions such as reading, creating, and deleting documents.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes documents and workflow data from the SBIS platform, which may contain content from untrusted third parties (e.g., invoices or messages from other companies).
  • Ingestion points: Document reading, list retrieval, and counterparty data access methods defined in SKILL.md.
  • Boundary markers: Absent; the instructions do not specify the use of delimiters or specific parsing rules to isolate external data from the AI's internal logic.
  • Capability inventory: The server includes powerful commands such as СБИС.УдалитьДокумент (Delete) and СБИС.УничтожитьДокумент (Destroy), as well as document signing capabilities.
  • Sanitization: Absent; there is no mention of filtering or sanitizing the content retrieved from the SBIS API before it is presented to the AI agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 10:51 AM
Security Audit — agent-trust-hub — sbis-mcp