- CollectionAgent Runtime & Infrastructure
- Classificationagent-native
- MCPUnavailable
- Agent SkillAvailable
- InterfacesURL onboarding · Agent Skill · REST · CLI · SDK · http · API
What it does
db9 is a serverless PostgreSQL platform designed for AI agents — combining a relational database with cloud-native filesystem storage in one unified workspace. Agents get SQL for structured state (metadata, embeddings, run history) and a filesystem for unstructured context (transcripts, reports, artifacts) — no S3 or external vector DB required.
Built-in superpowers: JSONB, pgvector-compatible vector search, HTTP calls from SQL, filesystem queries (fs9), cron jobs (pg_cron), full-text search with Chinese tokenizer, and environment branching that clones entire databases for isolated testing.
Agent-facing interfaces
- Interaction pattern: URL Onboarding ⭐ — the highest tier of agent-nativeness.
- Catalog protocols: URL onboarding, Agent Skill, REST, CLI, SDK, http, API
- MCP: ⚠️ Not yet published
- Agent Skill: ✅ URL Onboarding via skill.md — agent reads https://db9.ai/skill.md and follows instructions
- Onboarding type:
url
How agents onboard
Interaction pattern: URL Onboarding ⭐ — the highest tier of agent-nativeness.
Read https://db9.ai/skill.md and follow the instructions.
See the full onboarding section in the dossier.
Protocol surface
| Interface | Detail |
|---|---|
| CLI | db9 create, db9 db sql, db9 fs, db9 token create |
| REST API | https://api.db9.ai — databases, SQL execution, file ops |
| PostgreSQL | Direct pgwire to pg.db9.io:5433 |
| TypeScript SDK | get-db9 npm package — instantDatabase(), createDb9Client() |