Local-first observability dashboard for AI agents. MCP-native. Look at every span your agents emit.
Lookspan exposes only 1 tool (listTools) with a minimal schema. The tool name lacks a clear action verb (should be 'list_tool_calls' or similar). Description is present but generic (58 chars). Input schema exists with 2 parameters, both typed and described, but the schema is incomplete, no output schema is documented, no pagination guidance despite the tool accepting a 'limit' parameter. No error handling guidance. The tool appears to be a read-only observability query, but the response structure is not specified, making it impossible for an LLM to plan downstream actions or extract specific fields. This is a single-tool server with foundational gaps.
List recent tool calls across all traces with optional framework filtering
Tool name 'listTools' uses camelCase and lacks a clear action verb. Should be 'list_tool_calls' or 'list_recent_tool_calls' to match LLM expectations (verb_noun pattern).
No output schema documented. LLMs cannot infer what fields the response contains, making it impossible to chain this tool with others or extract specific data.
Tool accepts 'limit' parameter but provides no pagination guidance (no offset/cursor, no total count, no next_cursor). Large result sets will blow context windows.
No error handling guidance. If framework filter is invalid or no traces exist, the LLM has no recovery path.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 40 | 2026-07-28+ | v2 |
Description (58 chars) is below the 10 - 1024 char baseline and lacks context on WHEN to use this tool or WHAT it returns. Should explain: 'Returns recent tool invocations across traces, useful for debugging agent behavior or auditing tool usage.'