Documentation
guide/
- AGENTS.md Project-level instructions and context that load automatically into every session.
- Core Concepts
- Your first chain Walk from a blank file to a working authored chain in five edits.
- HITL Policies Control which tool calls require human approval using named policy presets.
- Quickstart Install Contenox and connect a model.
specification/
integrations/
- editors Use Contenox from AionUi Run your contenox chains inside AionUi — a free, local, open-source desktop chat UI for ACP agents.
- editors Use Contenox from JetBrains Drive your chains from inside GoLand, IntelliJ IDEA, and other JetBrains IDEs over the Agent Client Protocol.
- editors Use Contenox from OpenClaw Run Contenox as a hardened, contained ACP agent driven by OpenClaw over a chat channel — the untrusted-driver profile.
- editors Use Contenox from VS Code or VSCodium Install the Contenox extension from the Marketplace or a GitHub Release.
- editors Use Contenox from Zed Drive your chains from inside the Zed editor over the Agent Client Protocol.
- providers Anthropic Connect Contenox to Anthropic's Claude models via the Anthropic API.
- providers AWS Bedrock Connect Contenox to Amazon Bedrock — Claude, Llama, Mistral, and Nova models on your AWS account.
- providers Google Gemini Connect Contenox to Google Gemini via AI Studio or Vertex AI.
- providers Local Models (GGUF)
- providers Mistral Connect Contenox to Mistral's models via the Mistral API (La Plateforme).
- providers modeld The local inference daemon that powers Contenox's on-device models.
- providers modeld Architecture How modeld owns hardware, plans capacity, and manages sessions.
- providers Ollama Connect Contenox to a local Ollama instance or Ollama Cloud.
- providers OpenAI Connect Contenox to OpenAI or any OpenAI-compatible endpoint.
- providers OpenRouter Connect Contenox to OpenRouter's multi-provider model gateway.
- providers Vertex AI (Google Cloud) Configure Contenox to use Gemini on Vertex AI — billed through your GCP project — and renew credentials when they expire.
- What are Tools?
- tools Local Tools
- tools Model Context Protocol (MCP)
- tools Remote Tools
use-cases/
- Any API, a tool you authored Contenox turns any HTTP API into a scoped, credential-hidden, policy-governed tool — so an assistant, even an untrusted one, reaches exactly the slice you authored and nothing more.
- The pause is yours to define HITL isn't a checkbox. It's a policy file you wrote.
- Codebase Documentation
- Git & DevOps Recipes
- HubSpot via MCP
- Leads → HubSpot
- The moderation gate A small cheap model decides whether the big expensive one runs at all. Why I authored it this way.
- Multi-provider fallback as authored resilience The vendor doesn't choose your failure mode. You do.
- The nested permission bomb Why inheriting human permissions is a privilege escalation vulnerability, and how to author actual AI boundaries.
- Notion as a Tool
- Authoring your tool inventory Why partial OpenAPI registration is a game-changer for agent capabilities.
- Browser Automation with Playwright MCP Register Microsoft's Playwright MCP server and drive a real browser from the terminal — navigation, clicks, snapshots, and more.
- Automated Release Notes
- Stateful Agents with MCP
reference/
development/
- OpenAPI Generator Guide
- blueprints/acp OpenIDE Integration Blueprint
- blueprints/acp ACP Blueprints
- blueprints/acp/registry-submission ACP Registry Submission
- blueprints/acp Blueprint: Listing contenox on the Zed / ACP registry
- blueprints/beam OSS Beam auth — single operator, default `admin` / `admin`, changeable password
- blueprints/beam Blueprint: Beam Chat canvas
- blueprints/beam Plan: Revive HTTP API and Beam UI
- blueprints/beam Blueprint: Beam local runtime cockpit for modeld
- blueprints/beam Beam Blueprints
- blueprints/beam Blueprint: Beam Remote Connector
- blueprints Chat modes & context injection (Cursor-like)
- blueprints CI Release Hardening Blueprint
- blueprints Contenox Local Coding Node — Goals
- blueprints Contenox Local – Plan
- blueprints/modeld Plan: Alternative Silicon & AI PC Runtime Strategy
- blueprints/modeld Plan: lossless long context — derived hot/cold sizing (#5) + host-RAM KV cold store (#3)
- blueprints/modeld/effective-context Blueprint: modeld effective-context beyond the model's window
- blueprints/modeld/effective-context Blueprint: Backend Parity Contract
- blueprints/modeld/effective-context Blueprint: Benchmark Integrity and Reproducibility
- blueprints/modeld/effective-context Working Doc: Capacity/Context Bug Inventory (2026-07-02)
- blueprints/modeld/effective-context Blueprint: Latency-Budgeted Effective Context
- blueprints/modeld/effective-context Implementation Handover: T0 / T1 / T2 (backend-agnostic)
- blueprints/modeld/effective-context Local NVIDIA Llama Backend Facts
- blueprints/modeld/effective-context Blueprint: modeld Capability-Truth Boundary
- blueprints/modeld/effective-context Blueprint: modeld Local Inference Cross-Compare
- blueprints/modeld/effective-context Blueprint: Long Effective Context on a Consumer Accelerator
- blueprints/modeld/effective-context OpenVINO Backend Facts
- blueprints/modeld/effective-context OpenVINO/modeld Hardening Blueprint
- blueprints/modeld/effective-context Blueprint: Effective-Context Runtime Strategy
- blueprints/modeld/effective-context Blueprint: Session aac21f41 Learning Map
- blueprints/modeld/effective-context Blueprint: Specialization Cells and Multi-GPU Runtime Shapes
- blueprints/modeld Plan: modeld Interface Boundary — State vs. Compute
- blueprints/modeld/llama llama.cpp Binding Decision: Contenox-Owned Shim
- blueprints/modeld/llama Plan: Graduate Contenox Local Coding Node on llama.cpp
- blueprints/modeld/llama llama.cpp Llama Plumbing Log
- blueprints/modeld Blueprint: modeld LoRA adapters and local model variants
- blueprints/modeld Plan: Local Runtime Multi-Client Coordination
- blueprints/modeld/openvino Plan: Contenox Local Coding Node on OpenVINO
- blueprints/modeld/openvino OpenVINO Provider Plumbing — Handover
- blueprints/modeld/openvino OpenVINO S1.5 GenAI Provider Log
- blueprints/modeld/openvino OpenVINO S1 Embedded Controls Log
- blueprints/modeld/openvino OpenVINO S2.7 Parser Protocol Registry Log
- blueprints/modeld/openvino OpenVINO S2 Prefix-Reuse Proof Log
- blueprints/modeld Plan: ORT GenAI / Windows ML AI PC Runtime Track
- blueprints/modeld Plan: Local Runtime Owner — Lease Election and State Coordination
- blueprints/modeld Plan: modeld Provisioning and Detection
- blueprints/modeld modeld Blueprints
- blueprints/modeld Blueprint: modeld Release Artifacts
- blueprints/modeld Blueprint: modeld Setup Artifact Detection
- blueprints/modeld Plan: modeld Single Active Model Slot
- blueprints/modeld Blueprint: speculative execution for modeld
- blueprints/modeld Blueprint: TensorRT-LLM and the modeld Boundary
- blueprints/modeld Blueprint: Decouple modeld versioning; resolve by protocol compatibility
- blueprints Working Doc: Product Surface Audit — "does everything we surface actually work?"
- blueprints Blueprint: Product Surface Truth
- blueprints/providers Blueprint: AWS Bedrock provider
- blueprints/providers Plan: Cloud Provider Context Caching
- blueprints/providers Provider Blueprints
- blueprints Blueprints
- blueprints Remaining Work — local-node roadmap (post Phase 5)
- blueprints Contenox V1 Feature Map
- blueprints/vscode VS Code ACP-Shaped Permission Bridge Plan
- blueprints/vscode VS Code Bridge — Review Findings
- blueprints/vscode VS Code Extension — UX Shape Review
- blueprints/vscode Local Model Availability in Editor Surfaces
- blueprints/vscode VS Code Marketplace Release Blueprint
- blueprints/vscode VS Code Blueprints
- llama modeld Backend
- modeld Local Inference Landscape
- modeld Release Runbook
- modeld Source Build and Packaging