The Agent Skills Directory

[SAFE]: The skill implements standard JSON utilities (validation, formatting, querying, and diffing) using only the Python standard library. No external dependencies are required or installed, minimizing supply chain risks.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted JSON data from external sources, which creates a surface for indirect prompt injection if the processed content contains hidden instructions for the agent.
Ingestion points: Data is ingested via file paths, stdin, and command-line strings in scripts/validate_json.py, scripts/format_json.py, scripts/query_json.py, and scripts/diff_json.py.
Boundary markers: The scripts do not implement specific delimiters or 'ignore instructions' warnings when outputting data to the agent.
Capability inventory: scripts/format_json.py includes a file-write capability via the --output flag. All scripts output processed data to stdout for agent consumption.
Sanitization: The scripts do not sanitize or filter the content of the JSON fields; data is parsed and transformed according to the requested operation (querying, formatting) and returned as-is.

json-tools