docyrus-integrations-and-connectors
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill relies on the
docyrusCLI for all operations, including discovering integrations, listing connections, and executing actions. This is the intended functionality of the skill for terminal-based integration management. - [INDIRECT_PROMPT_INJECTION]: The skill enables the agent to fetch data from external, third-party services (such as Microsoft Graph, Twilio, and Meta) which are untrusted external sources.
- Ingestion points: Data is ingested through the output of
docyrus connect run-action,docyrus connect curl, anddocyrus curlcommands. - Boundary markers: The instructions do not define specific delimiters or "ignore" instructions to prevent the agent from following commands that might be embedded in the retrieved external data.
- Capability inventory: The agent can execute shell commands via the
docyrusCLI, which allows for network operations and management of third-party platform integrations. - Sanitization: Input parameters are validated against JSON schemas using AJV, but the skill does not specify sanitization or filtering for the data received back from external APIs.
Audit Metadata