amux server binary: HTTP API, SQLite store, orchestrator runtime, backends. A workspace orchestration platform with session management, file operations, browser automation, and task/workflow management.
18 tools with basic naming and descriptions, but critical gaps in schema documentation and parameter descriptions. Most tools lack visible input schemas in the source code provided. Descriptions are present but generic (10-50 chars typical), below the 50-200 char LLM-optimized baseline. No output schemas documented. Error handling and recovery guidance absent. Tools like set_prefs, update_schedule, approve_email, reject_email, create_board_task, update_board_task have minimal parameter documentation. No tool annotations (readOnlyHint/destructiveHint) despite clear read/write distinction in metadata.
Approve a pending email
Create a new task on the board
Get a specific task from the board
Get pending email approvals
Get pending grants
Get organization information including name and ID
Get user preferences and configuration settings
Get artifacts associated with a task
Input schemas missing or not visible for 12+ tools (health, get_org, get_prefs, list_sessions, list_connectors, list_schedules, get_email_approvals, get_grants, list_board_tasks, get_task_artifacts, get_task_verifications). Cannot verify parameter types, constraints, or validation rules.
Output schemas not documented for any tool. LLMs cannot infer what fields to expect, breaking downstream tool chaining and forcing agents to guess at response structure.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 46 | 2026-07-28+ | v2 |
Get verifications for a task
Server health check endpoint returning status, commit hash, and build information
List all tasks on the board
List all available connectors and their connection status
List all configured schedules and their execution information
List all available sessions with their status and metadata
Reject a pending email
Set user preferences and configuration settings
Update a task on the board
Update a schedule's enabled status
Descriptions are generic and under 50 characters for most tools (e.g., 'Get organization information' is 30 chars). LLM-optimized descriptions should be 50-200 chars and include WHEN to use the tool and what it returns.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite clear read/write distinction in metadata. Agents cannot determine which tools are safe to retry or have side effects.
No error handling or recovery guidance documented. Tools do not indicate what errors are retryable, user-fixable, or fatal, or provide actionable next steps for agents.