What is Coral?
Coral is an open source data retrieval layer that gives AI agents a unified SQL interface across APIs, databases, and file systems. It connects sources like GitHub, Slack, Datadog, and Jira into queryable tables and joins data across platforms in single queries, helping developers, SREs, and security teams build more accurate agents that reduce LLM costs by up to 70% compared to direct API integrations.
What sets Coral apart?
Coral sets itself apart with its local execution model that keeps all data queries and credentials within your own infrastructure, never sending sensitive information to external services. This approach proves particularly beneficial for security-conscious engineering teams and enterprises that need to maintain strict data governance while still accessing comprehensive context for incident response and debugging workflows. The open source architecture allows teams to audit query logic, customize source integrations, and retain full ownership of their AI-generated insights rather than being locked into vendor-controlled intelligence.
Coral Use Cases
- Cross-source data queries
- AI agent data retrieval
- Debugging and incident investigation
- Security investigations
- Customer support automation
Who uses Coral?
Features and Benefits
- Coral gives your agents a single SQL interface to query data across multiple APIs, databases, and file sources without writing separate integrations for each.
Unified SQL Interface
- Write a single SQL query that pulls and combines data from multiple sources at once, such as joining GitHub issues with Linear tickets in one statement.
Cross-Source Joins
- Coral exposes its query engine as an MCP server, so agents in Claude Code, Cursor, Codex, and other MCP-compatible clients can query your data sources directly.
MCP Integration
- Define per-source permissions and scoped credentials for your agents so they only access the data they need, keeping broad API credentials out of the LLM context.
Scoped Access Control
- Coral includes ready-to-use connectors for over 25 popular tools including GitHub, Slack, Datadog, Jira, Stripe, and PagerDuty, so you can start querying without building integrations.
Bundled Data Source Connectors
- Coral runs entirely on your own machine or cloud infrastructure, meaning your data, credentials, and query history never leave your environment.
Local Execution
Pricing
Coral query engine
Access control plane
Unlimited data sources
Unlimited agents
Unlimited queries
Deployment and support
Self-hosted
Use your own LLM keys
Community support
Everything in Open Source plus:
Admin tools
Social sign-in (Google, GitHub)
User roles and team filtering
Optimisation
Source connector evals
Deployment and support
Hosted on Coral
Automatic upgrades, no maintenance
Use your own LLM keys
Priority support via email and Slack
Everything in Team plus:
Self-managed options available
SAML and SCIM provisioning
Advanced RBAC and agent policy controls
Immutable audit log
Custom source connector development
Managed evals and optimisation
Flexible deployment (self-hosted, fully air-gapped or SaaS)
Uptime and support SLAs
Named support developers







