deepcontext-mcp

deepcontext-mcp is an MCP server that indexes a local codebase using AST-based parsing (tree-sitter) and builds a hybrid semantic search index (vector similarity + BM25 reranked with Jina). It exposes MCP tools to index code, search code, check indexing status, and clear indexes, enabling coding agents to retrieve more relevant code context than keyword search.

Evaluated Mar 30, 2026 (21d ago)
Homepage ↗ Repo ↗ DevTools mcp code-search semantic-search rag ast tree-sitter vector-search reranking developer-tools
⚙ Agent Friendliness
58
/ 100
Can an agent use this?
🔒 Security
54
/ 100
Is it safe for agents?
⚡ Reliability
29
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
70
Documentation
60
Error Messages
0
Auth Simplicity
85
Rate Limits
20

🔒 Security

TLS Enforcement
75
Auth Strength
55
Scope Granularity
20
Dep. Hygiene
55
Secret Handling
65

Uses API keys via environment variables (WILDCARD_API_KEY / TURBOPUFFER_API_KEY / JINA_API_KEY). The provided content does not describe secret logging/handling behavior, transport/TLS requirements explicitly, or whether tokens are scoped/least-privilege. The implementation likely relies on third-party embedding/reranking and vector storage services, implying code/derived data may be transmitted or stored externally in hosted mode; self-hosting shifts that responsibility but still uses external APIs unless fully local.

⚡ Reliability

Uptime/SLA
0
Version Stability
45
Breaking Changes
40
Error Recovery
30
AF Security Reliability

Best When

You are using MCP-capable coding agents and need fast, high-relevance retrieval of code symbols/chunks from large repositories, especially for mixed “keyword + semantic” queries.

Avoid When

You cannot run MCP servers (or cannot add MCP to your agent), or you require strict guarantees about how/where code and embeddings are stored/transmitted without documented controls.

Use Cases

  • Semantic symbol-aware code search for large TypeScript/Python repositories
  • Providing relevant code chunks to coding agents (Claude Code, Codex CLI, etc.) to reduce context window waste
  • Incremental reindexing as a codebase changes
  • Clearing and rebuilding an index when project structure or indexing configuration changes

Not For

  • General-purpose web search or non-code document retrieval
  • Applications that require a public REST/GraphQL API for querying (MCP is the primary interface)
  • Strict data residency/compliance needs that are not explicitly documented
  • Environments where outgoing calls to third-party embedding/reranking services are unacceptable (unless self-hosted)

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: Wildcard API key via WILDCARD_API_KEY (hosted backend)
OAuth: No Scopes: No

Auth is described as an API key environment variable for the Wildcard backend. Self-hosting uses TURBOPUFFER_API_KEY and JINA_API_KEY environment variables for external services. No fine-grained scopes are documented.

Pricing

Free tier: No
Requires CC: No

Pricing details are not provided in the supplied README content.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Long-running indexing is done via background processes; agents may need to poll get_indexing_status before relying on search results.
  • Indexing may exclude tests/generated/minified/build outputs; search results may not include those files.
  • Self-hosting requires API keys for vector storage and embeddings/reranking; hosted mode requires WILDCARD_API_KEY.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for deepcontext-mcp.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered