start-new-sdk-project

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill facilitates the use of the speakeasy CLI, instructing the agent to execute commands such as speakeasy quickstart and speakeasy run for project initialization and SDK generation.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external OpenAPI specifications from local paths or remote URLs, which constitutes an indirect prompt injection surface.
  • Ingestion points: The --schema flag in the speakeasy quickstart command (SKILL.md) allows loading data from external files and URLs.
  • Boundary markers: No explicit delimiters or isolation instructions are provided to separate schema data from agent instructions.
  • Capability inventory: Shell command execution and file system writes via the speakeasy CLI (SKILL.md).
  • Sanitization: The skill does not describe any sanitization or validation of the ingested schema content.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 07:46 PM
Security Audit — agent-trust-hub — start-new-sdk-project