// SCAN REPORT

CodeRAG

38 tools. 11 can modify or destroy data without limits.

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

Last updated: 11 Jun 2026

11 can modify or destroy data

27 read-only

38 tools total

Community server · catalogue entry verified 11/06/2026

How to control CodeRAG ↓

TOOL MAP

What CodeRAG exposes to your agents

Read (27) Write / Execute (8) Destructive / Financial (3)

Running more than one server? Check your whole stack →

MOST DANGEROUS TOOLS

Critical Risk

The most dangerous CodeRAG tools

Destructive · High clear_git_cache Clearing cached Git repositories is a destructive action because it permanently removes data (cached repository contents) and cannot be undone. Destructive · High delete_edge This tool permanently removes edges (relationships) from the Neo4J knowledge graph. Destructive · High delete_node This tool permanently removes nodes from the code knowledge graph, which cannot be easily undone. Write · Medium initialize_semantic_search Initializing semantic search infrastructure involves creating vector indexes and setting up data structures in the Neo4J database.

11 of CodeRAG's 38 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

POLICY

How to control CodeRAG

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

Deny destructive operations

{
  "clear_git_cache": {
    "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

{
  "initialize_semantic_search": {
    "limits": [
      {
        "counter": "initialize_semantic_search_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

{
  "analyze_testing_annotations": {
    "limits": [
      {
        "counter": "analyze_testing_annotations_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 CodeRAG — 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 CODERAG →

Free to start. No card required.

FULL CATALOGUE

All 38 CodeRAG tools

DESTRUCTIVE 3 tools

Destructive clear_git_cache Clear all cached Git repositories to free up disk space. Destructive delete_edge Delete an edge by ID Destructive delete_node Delete a code node by ID

WRITE 8 tools

Write initialize_semantic_search Initialize semantic search infrastructure including vector indexes Write scan_dir Scan a directory for source files and populate the graph Write scan_remote_repo Scan a remote Git repository and populate the graph database. Supports GitHub, GitLab, Bitbucket, and custom G Write add_edge Add a relationship edge between two nodes Write add_file Parse a single source file and add its entities to the graph Write add_node Add a new code node (class, interface, method, etc.) to the graph Write update_embeddings Generate or update semantic embeddings for code entities to enable semantic search Write update_node Update an existing code node

READ 27 tools

Read analyze_testing_annotations Analyze testing patterns and coverage based on test annotations/decorators Read calculate_ck_metrics Calculate Chidamber & Kemerer metrics for a class (WMC, DIT, NOC, CBO, RFC, LCOM) Read calculate_package_metrics Calculate package metrics (Afferent/Efferent Coupling, Instability, Abstractness, Distance) Read find_architectural_issues Find architectural issues (circular dependencies, god classes, high coupling) Read find_classes_calling_method Find all classes that call a specific method Read find_classes_implementing_interface Find all classes that implement a specific interface Read find_deprecated_code Find all code elements marked as deprecated and optionally their dependencies Read find_edges_by_source Find all edges originating from a source node Read find_implementations find_implementations Read find_inheritance_hierarchy find_inheritance_hierarchy Read find_method_callers find_method_callers Read find_nodes_by_annotation Find code nodes (classes, methods, etc.) that have specific annotations/decorators Read find_nodes_by_type Find nodes by their type Read find_untestable_code Find code patterns that may be difficult to test (private methods, static methods, etc.) Read find_usage_of_deprecated_code Find code that uses deprecated elements and assess migration impact Read get_annotation_usage Get comprehensive statistics on annotation/decorator usage patterns across the codebase Read get_edge Get an edge by ID Read get_framework_usage Get statistics on framework usage based on annotations/decorators across the codebase Read get_inheritance_hierarchy Get the inheritance hierarchy for a class Read get_node Get a code node by ID Read get_project_summary Get overall project metrics summary and quality assessment Read get_similar_code Find code entities that are semantically similar to a specific node Read git_cache_stats Get statistics about the Git repository cache, including number of cached repositories and total size. Read list_projects List all projects in the CodeRAG graph database with optional statistics Read search_nodes Search nodes by name, qualified name, or description Read semantic_search Search for code using natural language queries to find functionality by meaning rather than syntax Read validate_remote_repo Validate that a remote Git repository is accessible and can be scanned.

EXPLORE

Related servers

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

FAQ

Questions about CodeRAG

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

Yes. The CodeRAG server exposes 3 destructive tools including clear_git_cache, delete_edge, delete_node. 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 CodeRAG? +

The CodeRAG server has 8 write tools including initialize_semantic_search, scan_dir, scan_remote_repo. 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 CodeRAG.

How many tools does the CodeRAG MCP server expose? +

38 tools across 4 categories: Destructive, Execute, Read, Write. 27 are read-only. 11 can modify, create, or delete data.

How do I enforce a policy on CodeRAG? +

Register the CodeRAG 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 CodeRAG tool call.

Deterministic rules across all 38 CodeRAG tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

ENFORCE POLICY ON CODERAG →

Free to start. No card required.

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