Loading ToolBench
MCP server that gives AI agents full context about your multi-project ecosystem
ecosystem-mcp has 5 tools with basic descriptions and minimal input schemas. All tools are READ_ONLY and follow verb_noun naming (list_, get_, impact_). However, parameter descriptions are sparse or missing entirely. Most tools accept only a single string parameter with minimal constraint documentation. Output is unstructured text rather than typed JSON objects. No error handling guidance, no pagination support, and no documented output schemas. Tool descriptions are adequate (50-150 chars) but lack WHEN/WHY context. This is a typical community server with functional but incomplete definitions.
Show the dependency graph for a project — what it depends on and what depends on it.
Show shared infrastructure: databases, cache, domains, networking.
Get full info about a project: stack, dependencies, env vars, key files.
Impact analysis: what can break if this project changes?
List all projects in the ecosystem with status and stack.
No output schemas documented. All tools return unstructured text strings. LLMs cannot parse structured data or chain results to downstream tools.
Parameter descriptions missing or minimal. 'component' in get_infrastructure has a description, but 'name' in get_project and 'project' in get_dependencies lack context. LLMs cannot infer valid values.
No error handling guidance. Tools return bare 'not found' messages without recovery hints. LLMs cannot self-correct or know what to try next.
No pagination or result limits. list_projects returns all projects as a single text block. Large ecosystems will exhaust context windows.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 56 | 2026-07-28+ | v2 |
get_infrastructure accepts a free-form 'component' string. Should be an enum: ['databases', 'cache', 'domains', 'networking', 'all']. Prevents hallucinated values.