!
Purpose & Capability
The SKILL.md advertises full JMESPath/JSONPath support, syntax highlighting, and a CLI called json_tool.py; the bundle only contains scripts/json_formatter.py which implements a very small, custom query parser and no syntax highlighting or real JMESPath/JSONPath engine. The description overstates capabilities.
!
Instruction Scope
Runtime instructions reference json_tool.py and show examples (e.g., validate file.json) that do not match the provided script's name or its argument semantics. The script expects JSON text arguments, not filenames; following SKILL.md examples will produce errors. Instructions therefore are inconsistent and may cause unexpected failures.
✓
Install Mechanism
No install spec or external downloads are declared; the skill is instruction-only with a bundled local Python script. No remote fetches or installers are present.
✓
Credentials
The skill requests no environment variables, credentials, or config paths — consistent with a local JSON utility. There is no apparent attempt to access secrets or external services.
✓
Persistence & Privilege
always is false and the skill does not request elevated persistence or modify other skills. It does not attempt to persist credentials or alter agent-wide settings.
What to consider before installing
This package is internally inconsistent: the documentation and examples refer to a tool named json_tool.py and advanced features (JMESPath/JSONPath, syntax highlighting), but the repository only includes scripts/json_formatter.py, which provides basic JSON operations and a very limited query implementation. Before installing or invoking it: 1) don't run it with sensitive inputs or credentials (it doesn't request them, but the behavior is mismatched), 2) inspect and test the provided script in a sandbox with non-sensitive sample JSON to confirm behavior, 3) if you need full JMESPath/JSONPath, use a known library (e.g., jmespath) or a vetted tool, and 4) ask the publisher for clarification or a corrected SKILL.md and matching executable name. The code contains minor bugs/limitations (query and diff implementations are simplistic and may error), so treat it as experimental rather than production-ready.
