codevira-mcp 1.6.0

Persistent adaptive memory for AI coding agents — MCP server with context graph, semantic search, adaptive learning, roadmap tracking, and cross-tool continuity.

Codevira MCP

Persistent adaptive memory for AI coding agents — learns from every session, works with every tool, remembers across every project.

Works with: Claude Code · Claude Desktop · Cursor · Windsurf · Google Antigravity · any MCP-compatible AI tool

---

The Problem

Every time you start a new AI coding session, your agent starts from zero.

It re-reads files it has seen before. It re-discovers patterns already established. It makes decisions that contradict last week's decisions. It has no idea what phase the project is in, what's already been tried, or why certain files are off-limits.

You end up spending thousands of tokens on re-discovery — every single session.

Codevira fixes this.

---

What It Does

Codevira is a Model Context Protocol server that gives every AI agent working on your codebase a shared, persistent memory:

Capability	What It Means
Zero-config setup	Auto-detects language, source dirs, and file extensions; auto-injects IDE config — no prompts, no manual JSON editing
Live auto-watch	Background file watcher auto-reindexes on every save — no manual trigger needed
Context graph	Every source file has a node: role, rules, dependencies, stability, do_not_revert flags
Function-level call graph	Knows which function calls which — callers, callees, test coverage, risk scoring
Semantic code search	Natural language search across your codebase (optional — requires [search] extras)
Roadmap	Phase-based tracker so agents always know what phase you're in and what comes next
Changeset tracking	Multi-file changes tracked atomically; sessions resume cleanly after interruption
Decision log	Every session writes a structured log; past decisions are searchable by any future agent
Adaptive learning	Outcome tracking, confidence scoring, developer preference learning, and automatic rule inference
Cross-project memory	Learned preferences and rules sync across all your projects via ~/.codevira/global.db
Cross-tool continuity	Single "catch me up" call for seamless switching between Cursor, Claude Code, Windsurf, and Antigravity

The result: ~1,400 tokens of overhead per session instead of 15,000+ tokens of re-discovery.

---

Quick Start

1. Install

# Recommended: global install via pipx (isolated, works everywhere)
pipx install codevira-mcp

# Alternative: pip install
pip install codevira-mcp

# With semantic search support (adds ChromaDB + sentence-transformers)
pip install 'codevira-mcp[search]'

2. Register with your AI tools

cd your-project
codevira register

This single command — with zero prompts:

• Auto-detects language, source directories, and file extensions from project markers
• Auto-injects MCP config into Claude Code, Cursor, Windsurf, and Google Antigravity
• Installs a post-commit git hook for automatic reindexing

No explicit codevira init needed — the project is automatically initialized on the first MCP tool call (auto-init creates config, graph database, index, and roadmap on demand).

Note: codevira init is still available if you prefer to initialize upfront, but it is no longer a required step.

3. Verify

Ask your AI agent to call get_roadmap() — it should return your current phase and next action.

Note: After codevira init, restart your AI tool to pick up the new MCP server config.

Manual config (only if auto-inject didn't detect your tool)

Codevira supports two transports. Use the right one for your client:

Client	Transport	Config file
Claude Desktop (app)	stdio	~/Library/Application Support/Claude/claude_desktop_config.json
Claude Code (CLI)	stdio or HTTP	.claude/settings.json
Cursor	stdio	.cursor/mcp.json
Windsurf	stdio	.windsurf/mcp.json
Google Antigravity	stdio	~/.gemini/settings/mcp_config.json

Stdio transport — Claude Desktop, Cursor, Windsurf (.claude/settings.json / .cursor/mcp.json / .windsurf/mcp.json):

{
  "mcpServers": {
    "codevira": {
      "command": "codevira-mcp",
      "args": [],
      "cwd": "/path/to/your-project"
    }
  }
}

Claude Desktop (~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "codevira": {
      "command": "/path/to/codevira-mcp",
      "args": ["--project-dir", "/path/to/your-project"]
    }
  }
}

Tip: find the full binary path with which codevira-mcp

HTTP transport — Claude Code CLI via codevira-mcp serve (.claude/settings.json):

First start the HTTP server in a terminal:

codevira-mcp serve --port 7007 --project-dir /path/to/your-project
# For HTTPS (required by some clients):
codevira-mcp serve --https --port 7443 --project-dir /path/to/your-project

Then register the URL:

{
  "mcpServers": {
    "codevira": {
      "url": "https://localhost:7443/mcp"
    }
  }
}

HTTPS note: Claude Code uses Node.js, which requires a trusted CA for HTTPS. Run once to trust the mkcert CA:

brew install mkcert && mkcert -install
launchctl setenv NODE_EXTRA_CA_CERTS "$(mkcert -CAROOT)/rootCA.pem"
echo 'export NODE_EXTRA_CA_CERTS="$(mkcert -CAROOT)/rootCA.pem"' >> ~/.zshrc

Then restart Claude Code.

Auto-start on login (macOS):

codevira-mcp serve --install-service    # start server automatically on login
codevira-mcp serve --uninstall-service  # remove auto-start

Google Antigravity (~/.gemini/settings/mcp_config.json):

{
  "mcpServers": {
    "codevira-yourproject": {
      "$typeName": "exa.cascade_plugins_pb.CascadePluginCommandTemplate",
      "command": "codevira-mcp",
      "args": ["--project-dir", "/path/to/your-project"]
    }
  }
}

Codevira data layout (v1.6)

~/.codevira/                         <- global Codevira home
├── global.db                        <- cross-project intelligence
├── projects/
│   └── <project-key>/               <- per-project data (keyed by path)
│       ├── config.yaml
│       ├── metadata.json
│       ├── graph/
│       │   ├── graph.db
│       │   └── changesets/
│       ├── codeindex/               <- semantic search (optional)
│       └── logs/
└── certs/                           <- HTTPS certs (if using --https)

Legacy .codevira/ directories inside project repos are auto-migrated to centralized storage on first server start.

---

How It Works

Agent Session Lifecycle

flowchart TB

Start([Start Session])

subgraph Orientation
A[Check Open Changesets]
B[Get Project Roadmap]
C[Search Past Decisions]
D[Load Graph Context\nget_node - get_impact]
end

subgraph Execution
E[Plan Task]
F[Implement Code]
G[Run Tests / Validation]
end

subgraph Completion
H[Update Graph Metadata]
I[Write Session Log]
J[Complete Changeset]
end

Start --> A
A --> B
B --> C
C --> D
D --> E
E --> F
F --> G
G --> H
H --> I
I --> J

Code Intelligence Model

flowchart TB

A[Source Code]

subgraph Structural Analysis
B[Tree-sitter AST Parser]
C[Function / Class Extraction]
D[Dependency + Call Graph Analysis]
end

subgraph Knowledge Stores
E[(Semantic Index\nChromaDB — optional)]
F[(Context Graph + Call Graph\nSQLite DB)]
G[(Global Memory\n~/.codevira/global.db)]
end

subgraph Runtime Access
H[MCP Query Layer\n36 tools + 5 prompts]
end

I[AI Coding Agent\nClaude Code - Cursor - Windsurf - Antigravity]

A --> B
B --> C
C --> E

B --> D
D --> F

F --> H
E --> H
G --> H

H --> I

---

Session Protocol

Every agent session follows a simple protocol. Set it up once in your agent's system prompt — then your agents handle the rest.

Session start (mandatory):

list_open_changesets()      -> resume any unfinished work first
get_roadmap()               -> current phase, next action
search_decisions("topic")   -> check what's already been decided
get_node("src/service.py")  -> read rules before touching a file
get_impact("src/service.py") -> check blast radius

Session end (mandatory):

complete_changeset(id, decisions=[...])
update_node(file_path, changes)
update_next_action("what the next agent should do")
write_session_log(...)

This loop keeps every session fast, focused, and resumable.

---

36 MCP Tools + 5 Prompts

Graph Tools

Tool	Description
get_node(file_path)	Metadata, rules, connections, staleness for any file
get_impact(file_path)	BFS blast-radius — which files depend on this one
list_nodes(layer?, stability?, do_not_revert?)	Query nodes by attribute
add_node(file_path, role, type, ...)	Register a new file in the graph
update_node(file_path, changes)	Append rules, connections, key_functions
refresh_graph(file_paths?)	Auto-generate stubs for unregistered files
refresh_index(file_paths?)	Re-embed specific files in the search index
export_graph(format, scope?)	Export dependency graph as Mermaid or DOT diagram
get_graph_diff(base_ref?, head_ref?)	Show changed nodes, stability flags, and blast radius between git refs

Deep Graph Tools (v1.5)

Tool	Description
query_graph(file_path, symbol?, query_type)	Function-level: callers, callees, tests, dependents, symbols
analyze_changes(base_ref?, head_ref?)	Function-level risk scoring with test coverage gaps
find_hotspots(threshold?)	Large functions, high fan-in, high fan-out — complexity heatmap

Roadmap Tools

Tool	Description
get_roadmap()	Current phase, next action, open changesets
get_full_roadmap()	Complete history: all phases, decisions, deferred
get_phase(number)	Full details of any phase by number
update_next_action(text)	Set what the next agent should do
update_phase_status(status)	Mark phase in_progress / blocked
add_phase(phase, name, description, ...)	Queue new upcoming work
complete_phase(number, key_decisions)	Mark done, auto-advance to next
defer_phase(number, reason)	Move a phase to the deferred list

Changeset Tools

Tool	Description
list_open_changesets()	All in-progress changesets
start_changeset(id, description, files)	Open a multi-file changeset
complete_changeset(id, decisions)	Close and record decisions
update_changeset_progress(id, last_file, blocker?)	Mid-session checkpoint

Search Tools

Tool	Description
search_codebase(query, limit?)	Semantic search over source code (requires [search] extras)
search_decisions(query, limit?, session_id?)	Search all past session decisions
get_history(file_path)	All sessions that touched a file
write_session_log(...)	Write structured session record

Adaptive Learning Tools

Tool	Description
get_decision_confidence(file_path?, pattern?)	Outcome-based reliability scores
get_preferences(category?)	Learned developer style preferences
get_learned_rules(file_path?, category?)	Auto-generated rules from observed patterns
get_project_maturity()	0-100 intelligence score
get_session_context()	Single "catch me up" call for cross-tool continuity

Code Reader Tools

Tool	Description
get_signature(file_path)	All public symbols, signatures, line numbers (Python, TypeScript, Go, Rust)
get_code(file_path, symbol)	Full source of one function or class

Playbook Tool

Tool	Description
get_playbook(task_type)	Curated rules for: add_route, add_service, add_schema, debug_pipeline, commit, write_test

MCP Workflow Prompts (v1.5)

Prompt	Description
review_changes	Staged diff + blast radius + risk score
debug_issue	Symptom -> affected files -> call chain -> hypothesis
onboard_session	Full project context catch-up for new sessions
pre_commit_check	Test coverage gaps + high-risk functions before commit
architecture_overview	Module map + hotspots + dependency summary

---

Language Support

Feature	Python	TypeScript	Go	Rust	12+ Others
Context graph + blast radius	Y	Y	Y	Y	Y
Semantic code search	Y	Y	Y	Y	Y
Function-level call graph	Y	Y	Y	Y
get_signature / get_code	Y	Y	Y	Y
AST-based chunking	Y	Y	Y	Y
Auto-generated graph stubs	Y	Y	Y	Y
Roadmap + changesets	Y	Y	Y	Y	Y
Session logs + decision search	Y	Y	Y	Y	Y

Supported languages: Python, TypeScript, JavaScript, Go, Rust, Java, Kotlin, C#, Ruby, PHP, C, C++, Swift, Solidity, Vue.

---

Requirements

• Python 3.10+

Base install (pip install codevira-mcp): includes everything except semantic search. All 36 MCP tools work — graph, roadmap, changesets, code reader, learning, call graph.

With semantic search (pip install 'codevira-mcp[search]'): adds ChromaDB + sentence-transformers for search_codebase(). Downloads a ~90MB embedding model on first use.

---

Background

Want to understand the full story behind why this was built, the design decisions, what didn't work, and how it compares to other tools in the ecosystem?

Read the full write-up: How We Cut AI Coding Agent Token Usage by 92%

---

Contributing

Contributions are welcome. Read CONTRIBUTING.md for the full guide.

Reporting a bug? Open a bug report Requesting a feature? Open a feature request Found a security issue? Read SECURITY.md — please don't use public issues for vulnerabilities.

---

FAQ

Common questions about setup, usage, architecture, and troubleshooting — see FAQ.md.

Roadmap

See what's built, what's next, and the long-term vision — see ROADMAP.md.

License

MIT — free to use, modify, and distribute.