The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses npx to fetch the @pivanov/claude-wire package. This is a vendor-owned resource provided by the author to support the skill's functionality.
[COMMAND_EXECUTION]: Instructions direct the agent to execute the claude-wire CLI tool via Bash to process user prompts and JSON schemas.
[REMOTE_CODE_EXECUTION]: The use of npx involves downloading and executing remote code from a versioned npm package. This is the intended distribution method for the vendor's tool.
[PROMPT_INJECTION]: The skill processes untrusted input. 1. Ingestion points: User input passed via --prompt and data read from --schema-file or stdin. 2. Boundary markers: None explicitly defined. 3. Capability inventory: The tool performs an LLM call via the vendor's API but has no access to the local file system or network for tool use. 4. Sanitization: Output is strictly validated against a JSON schema.

ask-json