vaner 0.8.10

Predictive context middleware layer for AI workflows.

Vaner

Status: alpha (pre-1.0). Interfaces may evolve quickly while we stabilize core behavior.

What is Vaner?

Vaner is a local-first preparation engine for AI coding agents. It turns idle compute into evidence-backed Prepared Work — review notes, bug hypotheses, docs drift, virtual diffs, research briefs, ready prediction-backed drafts — and serves the best fit when the real question arrives.

Prepared Work is non-mutating by default: virtual diffs and exports require explicit user action. Vaner does not silently edit your files.

Install

Most users: download the desktop app at vaner.ai. The desktop app starts the engine, picks a local model for your hardware, and wires Vaner into the MCP clients it detects on your machine (Claude Code, Cursor, Zed, VS Code, Claude Desktop, …).

Power users / CI: install the CLI directly.

# Linux/macOS — one-line installer
curl -fsSL --proto '=https' --tlsv1.2 https://vaner.ai/install.sh | bash

# Or via your package manager of choice
pipx install 'vaner[mcp]'
uv tool install 'vaner[mcp]'

Full install matrix (flags, non-interactive CI, package upgrade, source build, retention/expiry): docs.vaner.ai/getting-started.

First run

vaner init --path .   # detect hardware, pick a model, wire detected MCP clients
vaner up --path .     # start the daemon + cockpit

The cockpit opens at http://127.0.0.1:8473/ and shows live pipeline state, prepared work, predictions, goals, and feedback.

Connect your AI client

Vaner exposes context to your agent over MCP. vaner init wires the clients it detects automatically. Manual setup:

Client	One command
Claude Code (plugin, recommended)	/plugin marketplace add Borgels/vaner then /plugin install vaner@vaner
Claude Code (manual MCP)	claude mcp add --transport stdio --scope user vaner -- vaner mcp --path .
Codex CLI	codex mcp add vaner -- vaner mcp --path .
Cursor, VS Code, Zed, Windsurf, Continue, Claude Desktop, Cline, Roo	see docs.vaner.ai/integrations

Documentation

Full docs at docs.vaner.ai:

• Getting started — install + first run
• Integrations — every supported MCP client
• Cockpit — live pipeline view + diagnostics
• Configuration — backends, compute, retention
• CLI reference — every command
• MCP tools — vaner.* namespace
• Architecture — how the engine works
• Security — local-first guarantees, threat model

Contributing

See CONTRIBUTING.md for dev setup, testing, CI, the Claude Code plugin parity rules, and DCO sign-off. Quick start:

git clone https://github.com/Borgels/vaner.git
cd vaner
python -m venv .venv && source .venv/bin/activate
pip install -e ".[dev]"
pre-commit install
pytest tests -m "not slow and not integration"

Community

• Security policy: SECURITY.md
• Governance: GOVERNANCE.md · maintainers: MAINTAINERS.md
• Code of conduct: CODE_OF_CONDUCT.md · support: SUPPORT.md
• Examples: examples/

License

Apache-2.0. Copyright 2026 Borgels Olsen Holding ApS (VAT DK39700425).