lark-markdown

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFECOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill relies on the lark-cli binary to perform all operations, executing shell commands to interact with the Lark/Feishu API for creating, updating, and fetching Markdown files.\n- [PRIVILEGE_ESCALATION]: The markdown +create shortcut implements a mechanism that, when used with the application identity (--as bot), automatically grants full_access permissions to the current CLI user for the newly created file.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes content from external Markdown files stored in the cloud, which could contain instructions designed to bypass agent constraints or influence its actions.\n
  • Ingestion points: External Markdown content is retrieved via the +fetch and +patch commands (referenced in references/lark-markdown-fetch.md and references/lark-markdown-patch.md).\n
  • Boundary markers: No delimiters or safety instructions are provided to the agent to treat the ingested Markdown data as untrusted content.\n
  • Capability inventory: The skill provides access to network resources (Lark Drive API), local file reading, and local file writing.\n
  • Sanitization: The skill does not perform validation or sanitization of the remote Markdown content before it is ingested into the agent's context.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 06:56 AM
Security Audit — agent-trust-hub — lark-markdown