twocaptcha-automation
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill establishes a workflow that ingests data from external tools and dynamic schemas, creating a potential surface for indirect prompt injection.
- Ingestion points: Results from RUBE_SEARCH_TOOLS and outputs from Twocaptcha toolkit operations as described in SKILL.md.
- Boundary markers: Absent; the instructions do not specify delimiters or warnings for the agent to ignore instructions embedded in external tool outputs.
- Capability inventory: High-privilege tool execution via RUBE_MULTI_EXECUTE_TOOL and RUBE_REMOTE_WORKBENCH.
- Sanitization: Absent; no explicit validation or filtering of content returned by the MCP tools is mentioned.
- [EXTERNAL_DOWNLOADS]: The skill configures a remote Model Context Protocol (MCP) server at https://rube.app/mcp for tool discovery and execution. This is a vendor-provided resource for the Composio ecosystem.
Audit Metadata