The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references an external installation guide hosted at https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md and provides instructions to install the belt-sh/cli utility. These resources originate from the inference.sh service and are part of the intended infrastructure for the skill's operations.
[DYNAMIC_EXECUTION]: The skill provides an example of using the infsh/python-executor tool to execute Python code for generating data visualization charts. The code uses the matplotlib library to create and save a PNG image. This is a standard feature for generating visual assets in report writing.
[INDIRECT_PROMPT_INJECTION]: The skill utilizes tools that ingest data from external search providers (Tavily and Exa). This creates a potential surface for indirect prompt injection if the retrieved content contains adversarial instructions designed to influence the agent's behavior.
Ingestion points: Results from Tavily and Exa search queries are processed within the agent context as described in SKILL.md.
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are present in the examples provided.
Capability inventory: The skill uses the Bash(belt *) tool, which allows executing various apps, including a Python runtime (infsh/python-executor).
Sanitization: There is no evidence of sanitization for external content before it is processed by the agent.

case-study-writing