The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the infsh CLI tool to execute commands for model selection and prompt submission. These operations are restricted to the infsh binary as specified in the allowed-tools configuration.
[EXTERNAL_DOWNLOADS]: The skill references installation instructions and additional model-related skills hosted on the inference-sh GitHub repository. These resources are directly related to the skill's primary function of providing model access.
[PROMPT_INJECTION]: The skill facilitates the ingestion of user-controlled prompts into the LLM inference workflow, which represents a standard attack surface for indirect prompt injection.
Ingestion points: User input is interpolated into the JSON payload of the infsh app run command within the SKILL.md examples.
Boundary markers: No explicit delimiters or boundary markers are utilized in the command-line examples to isolate the prompt content.
Capability inventory: The skill utilizes the infsh tool via the Bash shell.
Sanitization: No specific sanitization or filtering of the input prompt is performed by the skill instructions.

llm-models