openserv-agent-sdk

Pass

Audited by Gen Agent Trust Hub on Sep 22, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill provides patterns for building agents that ingest untrusted data from external sources and incorporate it into large language model prompts.
  • Ingestion points: Data enters the agent via trigger inputs (webhook, x402) and task data as demonstrated in SKILL.md, examples/basic-agent.ts, and examples/haiku-poet-agent.ts.
  • Boundary markers: The provided code examples lack explicit delimiters (e.g., XML tags or triple dashes) or instructions for the AI to ignore embedded commands within the interpolated data.
  • Capability inventory: Agents built with this SDK have access to workspace operations including uploadFile, deleteFile, createTask, and updateTaskStatus as seen in examples/file-operations.ts and examples/task-management.ts.
  • Sanitization: The templates do not demonstrate explicit validation or escaping of user-provided content before it is used in the prompt context.
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to download and install official vendor packages (@openserv-labs/sdk, @openserv-labs/client) and common dependencies from the npm registry.
  • [COMMAND_EXECUTION]: Documents the use of the platform's deployment CLI tool (npx @openserv-labs/client deploy) to package and upload agent code.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 22, 2026, 01:19 AM
Security Audit — agent-trust-hub — openserv-agent-sdk