# withcoral/coral

One SQL interface over APIs, files, and live sources — built for agents.

Repository: https://github.com/withcoral/coral
Canonical: https://ross.abutalabs.com/products/coral
Homepage: https://withcoral.com
Language: Rust
License: Apache-2.0
License Family: permissive
Last push: 2026-08-26T15:59:33+00:00

## Health v2 (maintenance only)
Score: 81/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 99, longevity 10
- inputs: {"age_days": 152, "days_push": 7, "days_rel": 8, "gap_med": 4.0, "n_releases_24m": 31}
- flags: young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4975, forks 215 (observed 2026-08-28T04:09:04.086153+00:00)

## What it is
Coral is a local SQL query engine that exposes APIs, files, and live data sources as read-only tables you can query and join with standard SQL. It ships as a desktop app and CLI, and exposes an MCP server so AI agents can retrieve data from multiple sources in fewer, more precise tool calls.

## Use cases
- query github slack and pagerduty data with sql joins from one interface
- give my ai agent a single sql tool instead of many mcp servers
- join api data across sources without writing pagination glue code
- reduce llm token costs when agents fetch company data
- run sql against live saas apis locally without sending data to the cloud
- connect claude code or cursor to my internal data sources over mcp

## When to choose
- you want agents to query multiple APIs or data sources with SQL instead of per-source MCP servers or API wrappers
- you need cross-source correlation (joins) in a single statement with tabular results
- you want a local-first tool where credentials and data never leave your machine
- you want to reduce agent tool-call count and token usage for read tasks

## When to avoid
- you need write access or mutations against your data sources, since Coral is read-only
- you need a fully managed hosted service on the free tier, since the open-source edition is self-hosted
- your queries are simple single-API fetches where a direct API call or existing MCP server suffices
- you need a source connector that has no bundled or community spec and you cannot write one

## Facets
- artifact type: application
- maturity: active
- function: database, mcp, agent-framework, api-framework, http-client, developer-tools, cli
- domain: artificial-intelligence, large-language-models, databases, developer-tools, apis
- platform: windows, cli, rust, cross-platform
- tags: sql-interface, federated-query, mcp-server, data-federation, agent-tools, local-first, source-connectors, ai-agents, data-engineering, automation, macos, linux, desktop

## Member repositories
- withcoral/coral (main) score 81

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:04.086153+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:17:44.456497+00:00, confidence not recorded.
  - readme: https://github.com/withcoral/coral (fetched 2026-08-28T04:09:04.086153+00:00, sha 91c4296519e6)
  - homepage: https://withcoral.com (fetched 2026-08-29T08:58:58.662466+00:00, sha 25f4b6e799c5)
  - site_page: https://withcoral.com/docs (fetched 2026-08-29T08:58:58.673398+00:00, sha d42eb69f8e40)
  - site_page: https://withcoral.com/pricing (fetched 2026-08-29T08:58:58.671729+00:00, sha 90ffd25bab6d)
- Data as of 2026-08-30T08:39:29.467469+00:00.