Function stringify

Format validation failure for LLM auto-correction feedback.

When LLM generates invalid function call arguments, this produces annotated JSON with inline // ❌ error comments at each invalid property. The output is wrapped in a markdown code block so that LLM can understand and correct its mistakes in the next turn.

Below is an example of the output format:

{
  "name": "John",
  "age": "twenty", // ❌ [{"path":"$input.age","expected":"number & Type<\"uint32\">"}]
  "email": "not-an-email", // ❌ [{"path":"$input.email","expected":"string & Format<\"email\">"}]
  "hobbies": "reading" // ❌ [{"path":"$input.hobbies","expected":"Array<string>"}]
}
Copy