Observal

A registry and insight platform for portable AI coding agents. Define context once, install it across tools, and learn what works.

If you find Observal useful, please consider giving it a star. It helps others discover the project and keeps development going.

What Observal is for

Observal is for teams doing context engineering across AI coding tools. If your organization maintains Skills, AGENTS files, MCP servers, hooks, prompts, sandboxes, or subagent definitions, Observal gives you one place to package them into versioned agents, publish them to a registry, and install them into the harness or harness your developers use.

Define an agent once. Observal renders the right configuration for Claude Code, Cursor, Kiro, Pi, Copilot, Codex, OpenCode, and other supported tools. As teams use those agents, Observal turns real usage into insights about which prompts, skills, tools, and policies are helping.

Why teams use Observal

• Package context into reusable agents: Bundle Skills, MCP servers, hooks, prompts, sandboxes, and policy into one versioned unit.
• Run a governed registry: Review submissions, approve internal agents, inspect version diffs, and give developers one trusted place to install from.
• Render across coding tools: Generate the correct config for each supported harness instead of maintaining separate setup instructions for every harness.
• Learn what works: Use real adoption and session data to find which agents, tools, prompts, and workflows are helping teams.
• Replay sessions when needed: Use traces as evidence for debugging, review, audits, and deeper analysis without making observability the main workflow.

Supported harnesses

One command to install any agent into any supported harness. The config files are generated per-harness automatically.

Quick Start

Observal has two parts: a server (API + web UI + databases) you self-host, and a CLI you install on each developer machine.

1. Deploy the server

One-line install (requires Docker Engine ≥ 24.0 with Compose v2):

curl -fsSL https://raw.githubusercontent.com/Observal/Observal/main/install-server.sh | bash

This downloads a Docker Compose package, runs guided setup (domain, secrets, ports), pulls container images from GHCR, and starts the full stack (API, web UI, PostgreSQL, ClickHouse, Redis, worker, load balancer, Prometheus, Grafana).

From source (for contributors):

git clone https://github.com/Observal/Observal.git && cd Observal
cp .env.example .env
make up

2. Install the CLI

Standalone binary (no Python required):

curl -fsSL https://raw.githubusercontent.com/Observal/Observal/main/install.sh | bash

Python (3.11+):

uv tool install observal-cli
# or: pipx install observal-cli

3. Connect your harness

observal auth login
observal doctor --patch

This authenticates with your server, detects your harness, installs telemetry hooks, starts capturing sessions automatically, and prepares it for agent installs and registry commands.

Once logged in, run /observal inside your harness and it takes the wheel. Pull agents, submit components, browse the registry, run diagnostics:

/observal pull security-auditor
/observal scan
/observal doctor

Or just tell your agent what you want and it figures out the right commands.

How Observal works

Agents are portable context packages

An agent bundles 5 component types into a single installable package: MCP servers, skills, hooks, prompts, and sandboxes. You define the agent once, publish it to the registry, and Observal generates the right config files for whichever supported harness or harness the user runs.

observal pull security-auditor --harness pi

The registry is the distribution layer

Browse published agents, see which harnesses they support, check download counts and ratings, and install with one command. Admins review submissions before they go live. Version diffs show exactly what changed between releases, so teams can safely evolve shared context.

Insights show what is helping

Observal turns real usage into reports about which agents, prompts, tools, and workflows are working or getting in the way. Use those insights to improve shared context instead of guessing from anecdotes.

Session traces provide the evidence

When you need to debug, audit, or understand a result, Observal can replay the full coding session: user prompts, thinking blocks, assistant responses, and tool calls with their inputs and outputs. The traces support registry and insight workflows rather than defining the product.

Agent Registry

Browse, search, and install agents with harness compatibility badges:

Build agents visually with live config preview for every harness:

Components library: MCPs, Skills, Hooks, Prompts, Sandboxes:

Agent Insights

AI-powered insight reports analyze usage patterns across all sessions, what's working, what's hindering, and quick wins. Powered by LiteLLM, works with any provider (Anthropic, OpenAI, Bedrock, Gemini, Azure, Ollama).

See Insights LLM Setup for configuration.

Session Replay

Full session overview with token counts, models, tools, and turn-by-turn timeline:

Every turn captured: user prompt, tool calls, thinking block, assistant response:

Drill into any span to see exact tool inputs and outputs:

Review and Governance

Admin review queue with full prompt inspection and approve/reject:

Version diffs show exactly what changed between releases:

Leaderboard tracks top agents and components by downloads:

Enterprise Edition

Source-available under a separate license. Activated with a signed JWT key. Core never imports from ee/, the open-source edition is fully functional without it.

Enterprise adds:

• Audit trail/logs with parameterized search and CSV export
• SAML SSO and SCIM provisioning
• Executive dashboard for org-wide agent performance

Audit log with parameterized search:

The server and CLI are the same package for all editions. Enterprise features activate at runtime when a valid