# dosco/graphjin

One governed graph for AI agents — GraphQL + MCP over your databases, files, APIs, and code

Repository: https://github.com/dosco/graphjin
Canonical: https://ross.abutalabs.com/products/graphjin
Homepage: https://graphjin.com
Language: Go
License: Apache-2.0
License Family: permissive
Topics: graphql, postgresql, sql, cloud-native, database, mysql, yugabyte, cockroachdb, mariadb, mongodb, mssql, oracle, sqlite, agentic-ai, ai-agents, llm, mcp
Last push: 2026-08-26T22:04:23+00:00

## Health v2 (maintenance only)
Score: 95/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 87, longevity 100
- inputs: {"age_days": 2719, "days_push": 7, "days_rel": 7, "gap_med": 0.0, "n_releases_24m": 157}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3150, forks 194 (observed 2026-08-28T04:07:46.143197+00:00)

## What it is
GraphJin is a Go compiler and runtime that turns GraphQL queries into optimized, policy-checked SQL across many databases, files, and APIs, exposed via GraphQL, REST/OpenAPI, and MCP. It doubles as a governed agent platform, letting AI agents discover, validate, and execute data operations with an auditable execution ledger instead of raw database credentials.

## Use cases
- expose my postgres database as a graphql api without writing backend code
- give ai agents safe governed access to company databases via mcp
- query multiple databases like postgres snowflake and mongodb through one graphql endpoint
- auto-generate a rest and graphql api from an existing sql schema
- let an llm answer business questions across databases files and saas apis safely
- build real-time graphql subscriptions over database changes
- prevent ai agents from running unsafe or unauthorized queries

## When to choose
- you want an instant, secure GraphQL/REST API over existing SQL databases with row and column-level policies
- you need a governed MCP/agent layer so LLMs can query and act on organizational data without raw credentials
- you need one graph spanning multiple databases, warehouses, and object stores with cross-database joins
- you want real-time subscriptions and an OpenAPI gateway from the same compiler

## When to avoid
- you need a fully custom hand-written GraphQL schema with arbitrary resolvers and third-party data sources
- your stack is not Go or container-friendly and you cannot run an external API service
- you only need a lightweight ORM inside application code rather than a standalone API layer
- you rely on databases GraphJin does not support

## Facets
- artifact type: framework
- maturity: active
- function: graphql, api-framework, mcp, agent-framework, database, http-server, rag
- domain: large-language-models, apis, databases, backend, web-development, developer-tools
- platform: go, windows, self-hosted, cli
- tags: graphql-to-sql, mcp-server, ai-agents, multi-database, governed-data-access, real-time-subscriptions, openapi-gateway, linux, macos, docker, web-server

## Member repositories
- dosco/graphjin (main) score 95

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:46.143197+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 2026-08-30T07:25:53.686865+00:00, confidence not recorded.
  - readme: https://github.com/dosco/graphjin (fetched 2026-08-28T04:07:46.143197+00:00, sha 79c015e716a3)
  - homepage: https://graphjin.com (fetched 2026-08-29T09:40:22.258461+00:00, sha eac080b9261e)
  - site_page: https://graphjin.com/start/install (fetched 2026-08-29T09:40:22.268010+00:00, sha 7930ab7f3098)
  - site_page: https://graphjin.com/integrations/multi-database (fetched 2026-08-29T09:40:22.269794+00:00, sha 17d7977c4011)
- Data as of 2026-08-30T08:39:29.467469+00:00.