Loading ToolBench
Git-backed Codex marketplace for the Fify information UI plugin.
Single tool 'render_information_ui' has a complete input schema with 7 parameters, all typed and described. However, the tool description (89 chars) is below the 194-char baseline and lacks critical context: WHEN to use it, WHAT it returns, and whether it modifies state. Parameter descriptions are present but generic (e.g., 'Array of content sections' lacks detail on required fields, nesting, or constraints). Output schema is not documented in the tool definition, only inferred from smoke test assertions. No error handling guidance. Naming is clear but the tool does one specific UI rendering task without composition context.
Renders information UI with structured content including sections, sources, and suggested refinements
Tool description is 89 characters, below 194-char baseline. Lacks WHEN to use, WHAT it returns, and state-modification intent. LLMs cannot determine selection criteria.
Output schema not documented in tool definition. Only inferred from smoke test (structuredContent.state, structuredContent.fallbackText, structuredContent.frames). LLMs cannot plan downstream calls without explicit return type documentation.
Parameter 'sections' description ('Array of content sections with id, title, body, items, and sourceIds') is generic. Does not specify: required vs optional fields, nesting depth, constraints on items array length, or valid sourceIds format.
No error handling guidance. Tool definition does not document failure modes (invalid locale, malformed sections, oversized payload) or recovery steps. Smoke test asserts no error but does not test error paths.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 52 | 2025-06-18+ | v2 |
Tool description does not state whether this is a read-only operation or if it has side effects. Risk field is 'READ_ONLY' but this is not reflected in the description text, which LLMs rely on.