The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user to install the belt CLI and additional technical skills using npx skills add, which fetches code from external repositories managed by the tool's ecosystem.
[COMMAND_EXECUTION]: The skill uses the belt tool (authorized in frontmatter) to execute various application tasks such as searching, image processing, and social media posting via belt app run commands.
[DYNAMIC_EXECUTION]: The skill demonstrates the use of infsh/python-executor to run Python code for generating charts, which involves dynamic execution of scripts in a managed environment.
[INDIRECT_PROMPT_INJECTION]: The skill defined workflow processes user-supplied topics and search results, creating a potential surface for indirect prompt injection.
Ingestion points: User topics and search queries are passed to belt app run exa/search in SKILL.md.
Boundary markers: No explicit delimiters or safety warnings are provided for interpolated user content in the provided examples.
Capability inventory: The skill has access to the belt CLI and the Python execution environment (infsh/python-executor) as described in SKILL.md.
Sanitization: No specific sanitization or filtering logic is described for handling external inputs before they are used in commands.

technical-blog-writing