// SCAN REPORT

Ensemble

19 tools. 6 can modify or destroy data without limits.

2 destructive tools with no built-in limits. Policy required.

Last updated: 30 Jun 2026

6 can modify or destroy data

13 read-only

19 tools total

Community server · catalogue entry verified 30/06/2026

How to control Ensemble ↓

TOOL MAP

What Ensemble exposes to your agents

Read (13) Write / Execute (4) Destructive / Financial (2)

Running more than one server? Check your whole stack →

MOST DANGEROUS TOOLS

Critical Risk

The most dangerous Ensemble tools

Destructive · High patterns_prune The tool irreversibly deletes data (patterns) from the vector memory system based on automatic criteria (age and match count). Destructive · High reset This tool irreversibly wipes all stored data in the ensemble-mcp system, including vector memory, indexed codebase, patterns, and session information. Write · Medium project_index The tool builds or refreshes an index, which is a write operation that creates or modifies stored index data. Write · Medium session_save This tool creates or modifies checkpoint state data for AI agent pipelines.

6 of Ensemble's 19 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

POLICY

How to control Ensemble

PolicyLayer is an MCP gateway — it sits between your AI agents and Ensemble, and nothing reaches the server without passing your rules. These are the rules we recommend:

Deny destructive operations

{
  "patterns_prune": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Blocked by default. Requires approval."
      }
    ]
  }
}

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations

{
  "patterns_store": {
    "limits": [
      {
        "counter": "patterns_store_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

{
  "context_compress": {
    "limits": [
      {
        "counter": "context_compress_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

Create a free account and register Ensemble — nothing to install.
Add these rules — paste them, or build them visually. Tune the limits to your setup.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

ENFORCE POLICY ON ENSEMBLE →

Instant setup, no code required.

FULL CATALOGUE

All 19 Ensemble tools

DESTRUCTIVE 2 tools

Destructive patterns_prune Remove old/unused patterns (zero match_count, older than max_age_days). Destructive reset Reset all stored data (destructive). Requires confirm=true.

WRITE 4 tools

Write patterns_store Store a new pattern from a successful pipeline for future semantic search. Write project_index Build or refresh the codebase index for faster Scope exploration. Write session_save Save pipeline checkpoint state with optimistic versioning. Write skills_generate Accept, dismiss, or defer a skill suggestion.

READ 13 tools

Read context_compress Compress verbose natural language text into terse, token-efficient form Read context_prepare Prepare and order prompt sections for optimal LLM cache hit rates. Read drift_check Check if code changes drift from the original task. Returns 0-1 drift score. Read health Server health check — status, version, DB size, pattern count. Read model_recommend Recommend a model tier (best/mid/cheapest) for an agent and task. Read patterns_search Search stored patterns by semantic similarity. Returns top-K matches. Read project_dependencies Get import/dependency graph for a specific file. Read project_query Query the project index — find files by type, path, or semantic query. Read project_snapshot Generate a compact project baseline summary from the codebase index. Read session_load Load latest or specific pipeline checkpoint. Read session_search Search sessions by semantic similarity. Returns top-K matches with scores. Read skills_discover Scan tool-native skill locations and return relevant skills via semantic search. Read skills_suggest Detect recurring patterns and suggest them as reusable skills.

EXPLORE

Related servers

Other MCP servers with similar tools — same risk classification, starter policies for each.

FAQ

Questions about Ensemble

Can an AI agent delete data through the Ensemble MCP server? +

Yes. The Ensemble server exposes 2 destructive tools including patterns_prune, reset. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through Ensemble? +

The Ensemble server has 4 write tools including patterns_store, project_index, session_save. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach Ensemble.

How many tools does the Ensemble MCP server expose? +

19 tools across 4 categories: Destructive, Execute, Read, Write. 13 are read-only. 6 can modify, create, or delete data.

How do I enforce a policy on Ensemble? +

Register the Ensemble MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Enforce policy on every Ensemble tool call.

Deterministic rules across all 19 Ensemble tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

ENFORCE POLICY ON ENSEMBLE →

Instant setup, no code required.

19 Ensemble tools catalogued and risk-classified — across an index of 43,000+ MCP servers.