Cairn

Cairn is a general-purpose problem-solving engine that searches for paths through unknown state spaces to reach a defined goal.

Cairn serves as a general-purpose engine for navigating unknown state spaces between a known origin and a defined goal. It utilizes a Blackboard Architecture with an explicit fact-intent graph to coordinate tasks without predefined roles. Agent Workers operate via an OODA loop—observing the graph, orienting to the state, deciding on intents, and acting to uncover facts. The system supports integration with Claude Code, Codex, and Pi models. It manages tasks through a server-dispatcher architecture that can execute jobs in isolated containers or directly on the host machine.

Cairn at a glance

Best for: Software engineer
Runs on: macOS, Linux
Pricing: Free under GNU AGPLv3 for personal and educational use; commercial licenses available upon request.
Released: April 2026
Category: AI coding tool
Website: github.com/oritera/Cairn

Key features

Blackboard-based task coordination
Explicit fact-intent graph structure
Autonomous OODA loop execution for agents
Containerized isolation per project
Support for Claude Code, Codex, and Pi backends
Local execution mode without Docker
Dynamic task generation at runtime

Use it for

Automated penetration testing
Security vulnerability research
Solving CTF challenges
Mathematical proof verification

Alternatives to Cairn

claude-mem — It shares persistent, searchable agent memory across sessions and devices.
ARTEX — Runs autonomous multi-agent penetration testing with a dual-graph architecture.
pstack — Improves AI coding agent outcomes with rigorous workflows.
Sandcastle — Orchestrates AI coding agents in isolated Docker, Podman, or Vercel sandboxes.
OpenViking — Unifies agent memory, knowledge RAG, and skills into a single directory structure.
OpenAPPA — Tracks data flows algebraically to prevent prompt injection without breaking agents.
jevgrep — Reduces coding agent costs by providing targeted, relevant repository context.
Ctxfw — Reduces AI context token mass while preserving code structure and dependencies.
Kiln — Combines a user-friendly desktop app with an open-source Python library.
Agent Beacon — Retains coding agent knowledge across tools so errors do not repeat.
Lumabri — Runs large AI models across multiple computers by sharing memory and layer workloads.
Metabob — Provides proactive debugging and enforces safe implementation patterns for AI tools.

All Cairn alternatives