The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of the belt CLI to run various applications for research and data processing.
Execution of belt app run for search assistants, browser agents, and data extraction tools.
Usage of belt login for authentication with the inference.sh platform.
[DYNAMIC_EXECUTION]: The skill includes a Python script executed via infsh/python-executor to generate a competitive positioning map.
The script utilizes matplotlib to render a 2D scatter plot and saves it as an image file (positioning-map.png).
This is a localized, low-risk use of dynamic execution for data visualization.
[EXTERNAL_DOWNLOADS]: The skill references external sources for tool installation and documentation.
Instructions to install the belt-sh/cli via npx.
A link to a CLI installation markdown file hosted on the inference-sh GitHub repository.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external, untrusted sources which presents a potential injection surface.
Ingestion points: Data is fetched from search results (tavily/search-assistant, exa/search) and direct URL extraction (tavily/extract).
Boundary markers: None identified in the provided templates to delimit untrusted search results from agent instructions.
Capability inventory: The skill has access to python-executor, agent-browser (screenshot capability), and local file-writing through stitch-images and the python script.
Sanitization: There are no explicit instructions for the agent to sanitize or filter the content retrieved from competitor websites or search engines before including it in final reports.

competitor-teardown