The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches installation instructions and configuration from a remote repository.
Evidence: The skill points to https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md for setup instructions.
[REMOTE_CODE_EXECUTION]: The skill instructs the agent to use a Python execution tool to generate visual content like benchmark charts.
Evidence: Examples in SKILL.md demonstrate using belt app run infsh/python-executor to execute Python code provided as a string.
[COMMAND_EXECUTION]: The skill requests permission to execute shell commands specifically limited to the belt CLI.
Evidence: allowed-tools: Bash(belt *) in the YAML frontmatter.
[PROMPT_INJECTION]: The skill processes technical research and user topics to generate content, which presents a surface for indirect prompt injection if the inputs contain malicious instructions.
Ingestion points: Ingests user-provided research data, topic queries, and code snippets during the blog writing process.
Boundary markers: None present; the instructions do not specify delimiters to separate untrusted data from instructions.
Capability inventory: The skill has the ability to execute shell commands (belt), perform network searches (exa/search), and execute Python code (python-executor).
Sanitization: No sanitization or validation of the input content is described before it is processed by the tools.

technical-blog-writing