No user-submitted related servers found.
Scored across 43 tools
The descriptions are unusually detailed and explicitly disambiguate many neighboring tools, but the set still contains overlapping clusters such as create_task vs start_project_request, dispatch_agent vs advance_project, resume_pending_dispatch vs recover_workers, and several read-only status/context/evaluation readers. An agent can generally tell them apart by reading the long descriptions, but the cognitive load is high and boundaries are not obvious from names alone.
Almost every tool follows a predictable snake_case verb_noun pattern, e.g. list_projects, get_task_status, propose_context_patch, submit_agent_report. The few compound names remain readable and consistent within the same convention.
At 43 tools, the surface is far beyond the typical 3-15 well-scoped range and beyond the 25+ threshold for being too many. Many tools are meaningful in this orchestration domain, but the sheer size makes selection and maintenance burdensome.
The tool set covers the project lifecycle comprehensively: project discovery/bootstrap/creation, task start/status/wait/advance/cancel/recovery, context read/validate/patch/commit/history, workspace listing/reading/writing, approvals, reviews, evaluations, reliability, meeting clarifications, and installed MCP tool access. Read-only settings and local approval are intentional boundaries rather than clear gaps.