A
licenseNot graded
qualityB
maintenanceMCP server that lets Claude Code or any MCP client delegate coding tasks to the local Cursor Agent CLI, with persistent per-workspace sessions and tools for diffing and pushing changes.
12 npm
MIT
ModelBound
Officialby ModelBound
Related Servers
Alternatives to ModelBound
No user-submitted related servers found.
Related Servers
- AlicenseAqualityAmaintenanceA universal MCP server that enables any LLM or AI agent to access expert skills from your local filesystem.3112 npm40MIT
- AlicenseNot gradedqualityDmaintenanceA management toolkit for AI agent skills that provides an MCP server for search-first skill discovery and on-demand loading. It enables users to validate, organize, and serve standardized skills to MCP-compatible clients like Cursor and GitHub Copilot.567 PyPI414MIT
- AlicenseNot gradedqualityDmaintenanceMCP server that adapts, activates, and shares local Agent Skills for runtimes like OpenAI Agents SDK and Claude Code, enabling dynamic discovery and shareable HTML/image artifacts.13 npm2MIT
- AlicenseNot gradedqualityCmaintenanceLocal-first MCP server that provides project context, verification gates, and structured tools for coding agents to discover knowledge, run diagnostics, and execute allowlisted commands within a repository.33 npmMIT
- AlicenseBqualityCmaintenanceMCP (Model Context Protocol) Server for skill4agent - Search, view, and install AI skills in AI conversations.319 npm4MIT
TDQS
C2.9/5.0
Scored across 51 tools
Disambiguation3/5
Most tools target distinct actions, but several check/safety tools overlap heavily (skills.checks, skills.flightCheck, check_harness, skill.findings all 'check a skill' with subtly different scopes), and test tools (skill.test, eval.run, skill.benchmark, skill.testRuns) blur together. Detailed descriptions help, but selection still requires careful reading.
Naming Consistency2/5
Mix of dot-namespaced and flat snake_case tools (check_harness, report_run) with no unified scheme, and even the namespaces are inconsistent (skills. vs skill.). Method casing is also mixed, e.g. skill.report_outcome (snake) alongside skills.reviewRequest (camelCase).
Tool Count2/5
51 tools is heavy for a skill-management server and clearly exceeds the 25+ 'too many' threshold. Multiple families (review, checks, tests, optimization) contain near-duplicate capabilities that inflate the surface beyond what the core purpose requires.
Completeness4/5
Broad lifecycle coverage: scaffold/create, read/write, cloud push/pull/list, versioning/diff, review workflow, testing, eval, optimization, and outcome reporting are all present. Only minor gaps remain, and agents can work around them.
Maintenance
ActivityMaintained
ResponsivenessNo issues