Skip to main content
Glama
umuterdal

Blindspot MCP

by umuterdal
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Local

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
PROJECT_PATHNoThe path to the project directory. Can also be set via --project-path argument.

Instructions

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

This server publishes no instructions, or was last inspected before Glama recorded them.

Capabilities

Features and capabilities supported by this server

Protocol revision2025-11-25

CapabilityDetails
tools
{
  "listChanged": false
}
prompts
{
  "listChanged": false
}
resources
{
  "subscribe": false,
  "listChanged": false
}
experimental
{}

Tools

Functions exposed to the LLM to take actions

NameDescription
set_project_pathB

Call this first. Sets the active repo and initializes the available index state.

search_codeA

Fallback text search. Use this only when get_context and symbol tools are not enough.

find_filesB

Locate likely files by glob pattern, then use get_context on the best candidate.

get_symbol_bodyA

Return one symbol's metadata or bounded source. Use this when exact symbol code is needed.

refresh_indexA

Rebuild the shallow file index when file discovery looks stale.

build_deep_indexB

Build the deep symbol index required for rich relationship and impact analysis.

get_edit_regionC

Return a tight numbered excerpt around a symbol or line range for precise edits.

get_project_snapshotA

Session-start overview of the repo. Use this once before drilling into a specific edit.

get_contextA

Primary tool. Returns normalized file, symbol, relationship, and impact context before editing.

Pass owner (for example the enclosing class) when a symbol name is ambiguous within a file, so the engine can disambiguate identically named methods such as User.save vs Order.save.

Pass query as a free-text description (e.g. "pricing quote total") when you do not know the exact file or symbol. The engine runs a BM25 search over the indexed symbols, resolves target/symbol/owner from the top match, and echoes the inference under query_resolution.

If target points to a previously-saved detail file under .blindspot/output/*.json, the stored payload is returned verbatim so oversized responses never strand the caller.

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources

Deploy Server
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

A3.7/5.0

Scored across 9 tools

Disambiguation5/5

Each tool has a distinct purpose: initialization, index building, file discovery, context retrieval, symbol extraction, edit region extraction, overview, search fallback, and path setting. No two tools appear to do the same thing, minimizing ambiguity.

Naming Consistency5/5

All tool names follow a consistent verb_noun pattern with snake_case (e.g., build_deep_index, get_context, set_project_path). No mixing of naming conventions, which aids predictability.

Tool Count5/5

With 9 tools, the server covers essential operations for code analysis and edit preparation without being bloated. Each tool earns its place, and the scope is well-defined.

Completeness4/5

The tool set covers project initialization, index management, file discovery, context retrieval, symbol detail, edit regions, and project overview. Minor gaps like a path-switching tool or direct edit application are absent, but the primary focus on analysis and context is well-served.

Maintenance

ActivityInactive
ResponsivenessNo issues

Resources