ollama
Pass
Audited by Gen Agent Trust Hub on Apr 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill facilitates legitimate management of local AI models using Ollama. All activities described are consistent with the tool's intended use.\n- [COMMAND_EXECUTION]: The skill uses the
xCLI to perform model management tasks such as pulling models from the registry (x ollama pull) and accessing an interactive model browser (x ollama la). These actions are standard for the tool's operation.\n- [PROMPT_INJECTION]: The skill enables processing of local files as context for the LLM using the--fileflag, which is a standard ingestion point for external data.\n - Ingestion points: Local file content passed to the model via command-line arguments (e.g., abstract.en.md in SKILL.md).\n
- Boundary markers: None explicitly specified in the instructions; the agent is directed to use the file content as context directly.\n
- Capability inventory: Interacting with local models for chat, registry synchronization (pull/push), and file-based translation tasks.\n
- Sanitization: No explicit sanitization or validation of input file content is documented in the skill instructions.
Audit Metadata