Skip to content
Back to search

deepcontext-mcp

DeepContext is an MCP server that adds symbol-aware semantic search to Claude Code, Codex CLI, and other agents for faster, smarter context on large codebases.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/wildcard-official--deepcontext-mcp/badge.svg)](/listings/wildcard-official--deepcontext-mcp)

HTML

<a href="/listings/wildcard-official--deepcontext-mcp"><img src="/api/listings/wildcard-official--deepcontext-mcp/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

1. Visit the Wildcard [DeepContext](https://wild-card.ai/deepcontext) page 2. Click "Generate API Key" 3. Copy your API key 4. Paste installation command for your MCP client 5. Type `index this codebase` to index the current directory

**Claude Code:**

claude mcp add deepcontext \
  -e WILDCARD_API_KEY=your-wildcard-api-key \
  -- npx @wildcard-ai/deepcontext@latest

**Codex:**

# Add to ~/.codex/config.toml
[mcp_servers.deepcontext]
command = "npx"
args = ["-y", "@wildcard-ai/deepcontext@latest"]
env = { "WILDCARD_API_KEY" = "your-wildcard-api-key" }

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 151 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. <p align="center">
  2. <img alt="DeepContext Logo" src="https://github.com/user-attachments/assets/6dd1b4a4-5ce3-41cb-acc0-b3eed32d5f8e" width="500" height="296">
  3. </p>
  4. <h1 align="center">Smarter Context for Large Codebases</h1>
  5. <p align="center">
  6. <a href="https://github.com/Wildcard-Official/deepcontext-mcp/stargazers">
  7. <img src="https://img.shields.io/github/stars/Wildcard-Official/deepcontext-mcp?style=social" alt="GitHub stars">
  8. </a>
  9. <a href="https://www.npmjs.com/package/@wildcard-ai/deepcontext">
  10. <img src="https://img.shields.io/npm/v/@wildcard-ai/deepcontext" alt="NPM Version">
  11. </a>
  12. <a href="https://join.slack.com/t/wildcard-community/shared_invite/zt-3dslre0yd-H8M_Ftsf8AJoB6nP~wVIpg">
  13. <img src="https://img.shields.io/badge/Slack-join-blueviolet?logo=slack" alt="Slack Community">
  14. </a>
  15. <a href="https://x.com/wildcard_ai">
  16. <img src="https://img.shields.io/twitter/follow/wildcard_ai?style=social" alt="Twitter Follow">
  17. </a>
  18. </p>
  19. DeepContext is an MCP server that adds symbol-aware semantic search to Codex CLI, Claude Code, and other agents, giving them more precise context of even the largest codebases. Currently supports Typescript and Python.
  20. ## Quickstart
  21. 1. Visit the Wildcard [DeepContext](https://wild-card.ai/deepcontext) page
  22. 2. Click "Generate API Key"
  23. 3. Copy your API key
  24. 4. Paste installation command for your MCP client
  25. 5. Type `index this codebase` to index the current directory
  26. **Claude Code:**
  27. ```bash
  28. claude mcp add deepcontext \
  29. -e WILDCARD_API_KEY=your-wildcard-api-key \
  30. -- npx @wildcard-ai/deepcontext@latest
  31. ```
  32. **Codex:**
  33. ```toml
  34. # Add to ~/.codex/config.toml
  35. [mcp_servers.deepcontext]
  36. command = "npx"
  37. args = ["-y", "@wildcard-ai/deepcontext@latest"]
  38. env = { "WILDCARD_API_KEY" = "your-wildcard-api-key" }
  39. ```
  40. ## Demo
  41. https://github.com/user-attachments/assets/9a2d418f-497b-42b9-bbb2-f875ef0007b4
  42. ## Why DeepContext MCP?
  43. Most coding agents use grep based search that match exact text, these searches miss semantically related code and fill context windows with irrelevant results. Large codebases amplify this problem, where text search returns hundreds of matches that quickly overwhelm conversation capacity. This leads to slow completions, more hallucinations, and lower success rates.
  44. DeepContext provides agents with intelligent search that preserves context windows by finding only relevant code chunks.
  45. - **Semantic accuracy**: Matches code by meaning and relationships rather than text patterns, finding related functions across files that keyword search misses.
  46. - **Reduced token usage**: Returns precise code chunks instead of every file containing your search terms, preserving conversation context windows and reducing costs.
  47. - **Search speed**: Searches code immediately through pre-indexed data for instant file discovery.
  48. ## MCP Tools
  49. #### `index_codebase`
  50. Creates a searchable index of your codebase for semantic search.
  51. #### `search_codebase`
  52. Finds relevant code using natural language or keyword queries.
  53. #### `get_indexing_status`
  54. Shows indexing status and file counts for your codebases.
  55. #### `clear_index`
  56. Removes all indexed data for a codebase.
  57. ## Architecture
  58. **MCP Integration Flow**
  59. - Coding Agent communicates with DeepContext through the Model Context Protocol
  60. - MCP server receives requests, validates parameters, and routes to appropriate core components
  61. - For long-running operations like indexing, spawns detached background processes to prevent timeouts
  62. - Background workers handle large codebases without blocking MCP channel // Reword
  63. **AST-Based Parsing**
  64. - Tree-sitter parsers analyze source code to build Abstract Syntax Trees
  65. - Python, TypeScript, and JavaScript language grammars for accurate parsing
  66. - Semantic node identification for functions, classes, interfaces, and modules
  67. - Symbol extraction identifies functions, classes, interfaces, types, variables, and constants
  68. - Scope analysis determines local vs exported vs global visibility
  69. - Parameter and return type extraction for function signatures
  70. - Import/export analysis maps module dependencies and cross-file relationships
  71. - Creates chunks at semantic boundaries rather than arbitrary line or token splits
  72. - Large file handling through range-based parsing with overlapping windows
  73. **Hybrid Search with Reranking**
  74. - Search operates in three stages
  75. - Hybrid search combines vector similarity and BM25 full-text search
  76. - Jina reranker-v2 for final relevance optimization
  77. - Vector similarity finds semantically related code using embeddings
  78. - Jina text embeddings generate 1024-dimension vectors for code chunks
  79. - BM25 performs traditional keyword matching for exact terms
  80. - Full-text indexing enables precise identifier and comment matching
  81. - Results fused using configurable weights, then reordered by Jina reranker
  82. **Incremental Indexing**
  83. - Uses file modification times and content hashes to track changes
  84. - SHA-256 hashing detects content modifications at byte level
  85. - Only reprocesses files with different hashes during reindexing
  86. - Avoids unnecessary parsing and embedding generation for unchanged files
  87. **Content Filtering**
  88. - Scores files based on extension patterns, path components, and content analysis
  89. - Language detection and file type classification for processing decisions
  90. - Excludes test files, generated code, minified files, and build outputs during indexing
  91. - Pattern matching against common test frameworks and build tool outputs
  92. - Filters documentation and configuration files to focus on source code
  93. ## Self Hosting
  94. Self-hosting requires code modifications to integrate directly with vector storage and embedding providers, as the current implementation uses the Wildcard API backend.
  95. **Prerequisites**
  96. - Node.js 20+ for ES module support and performance optimizations
  97. - Turbopuffer API key for vector storage and hybrid search operations
  98. - Jina AI API key for text embeddings and reranking services
  99. **Setup**
  100. ```bash
  101. git clone https://github.com/Wildcard-Official/deepcontext-mcp.git
  102. cd deepcontext
  103. npm install
  104. npm run build
  105. ```
  106. **Integration**
  107. ```bash
  108. claude mcp add deepcontext-local \
  109. -e TURBOPUFFER_API_KEY=your-turbopuffer-key \
  110. -e JINA_API_KEY=your-jina-key \
  111. -- node /path/to/deepcontext/dist/standalone-mcp-integration.js
  112. ```
  113. ## Contributing
  114. Thanks for your interest! We’re currently not accepting external contributions as we’re an early-stage startup focused on rapid iteration. We may open things up in the future — feel free to ⭐ the repo to stay in the loop.
  115. ## License
  116. Licensed under the Apache License.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.