Agentset
Open-source RAG platform with built-in citations, deep research, 22+ file formats, MCP server, and more.
- Source repo
- agentset-ai/agentset
- Stars
- ★ 2.1k
- Last updated
- 20d ago
- License
- MIT
- Primary language
- TypeScript
- FA score
- 33/100 · Major gaps
At a glance
- How it runs
- Works with
- Universal · cross-platform
- Cost
- Free tier plus a paid hosted plan
- Setup effort
- Medium · a few setup steps
- You'll need
- Typical use
- A developer wants to quickly set up a RAG prototype and deploy it to the cloud without managing infrastructure.
- Not a fit if
- Teams that need a fully offline setup without Supabase or other cloud services
- Teams unwilling to adopt the Next.js, Prisma, and Trigger.dev stack
- Teams that cannot use Bun as their package manager
- Source review
- 33/100 · Major gaps 8 safety controls not found
What does this agent do, and when should you use it?
Agentset is an open-source platform for building, evaluating, and shipping production-ready RAG and agentic applications. It provides end-to-end tooling: ingestion, vector indexing, evaluation/benchmarks, a chat playground, hosting, and a developer-friendly API. It is model agnostic, working with your choice of LLM, embeddings, and vector DB. Built-in multi-tenancy and typed SDKs are included. The stack is TypeScript, Next.js, AI SDK, Prisma, Supabase, and Trigger.dev.
Agentset powers RAG pipelines with ingestion, chunking, embeddings, and retrieval. It supports 22+ file formats and offers a chat playground with message editing and citations. It handles production hosting with preview links and custom domains. Developers interact via a REST API with typed SDKs generated from OpenAPI spec. It includes built-in multi-tenancy. Administration uses Prisma for database migrations, and the frontend runs on Next.js; background jobs leverage Trigger.dev.
- A developer wants to quickly set up a RAG prototype and deploy it to the cloud without managing infrastructure.
- A team needs a chat interface with citations for a knowledge base, using custom LLM, embeddings, and vector DB.
- A SaaS provider needs multi-tenant RAG capabilities for multiple customers.
- A developer wants to integrate RAG into an existing application using the provided API and SDKs.
- An organization wants to self-host a RAG platform to keep data on-premises, following the self-hosting guide.
How do you install or deploy this agent?
For local development: copy .env.example to .env and fill required values. Install dependencies with bun install. Run database migrations with bun db:deploy. Start the app with bun dev:web. Alternatively, use Agentset Cloud for a hosted version without self-hosting.
How do you use this agent?
- Configure your environment variables. 2. Run database migrations and start the web app. 3. Configure your LLM, embeddings, and vector DB in the chat playground. 4. Ingest documents and use retrieval with citations. 5. Integrate with your app via API/SDK. Useful scripts:
bun db:studiofor Prisma Studio,bun dev:webto run only the web app.
What are this agent's strengths and limitations?
- Model agnostic: supports your choice of LLM, embeddings, and vector DB.
- Built-in citations and deep research features enhance explainability.
- Both cloud and self-hosted deployment options provide flexibility.
- Multi-tenancy built-in, suitable for SaaS scenarios.
- Typed SDKs and OpenAPI specification simplify integration.
- Tied to a specific stack: Next.js, Prisma, Supabase, Trigger.dev, requiring familiarity.
- Self-hosting requires managing your own database and infrastructure.
- Documentation may not cover all configuration options in depth.
- Requires Bun as a package manager, adding a learning curve.
How does this agent compare with similar options?
Key facts side by side with the most closely related agents.
| Agent | Source review | Form / cost | Stars | Updated | Language | Full support on |
|---|---|---|---|---|---|---|
| Agentset This agent | 33 · Major gaps | Hosted serviceFreemium | ★ 2.1k | 20d ago | TypeScript | — |
| LangChain | 41 · Major gaps | Library / SDKFree + model costs | ★ 148k | today | Python | OpenAI API |
| Full-Stack AI Agent Template | 74 · Some gaps | CLIFree + model costs | ★ 2k | 12d ago | Python | OpenAI API · Claude API |
| Skybridge MCP Apps Framework | 64 · Some gaps | CLIFreemium | ★ 2.2k | 4d ago | TypeScript | ChatGPT · Claude.ai |
How does FollowAgents rate this agent?
Why each dimension lost points
Evidence shows an open-source RAG platform, but no specific implementation of least privilege, user confirmation, data flow transparency, sensitive data handling, dependency security, external effects, rollback, or source attribution. Hence all trust criteria scored 0.
Test files show detailed tests for system prompt resolution and Pinecone filter translation, indicating good internal consistency. However, dependency availability is only declared in package.json, not verified. Failure messages are not explicitly documented.
README describes multiple usage scenarios (cloud, self-hosted, local dev), but capability boundaries are not explicit. Trigger precision is not detailed. Environment fit is addressed with Node.js and Bun requirements, but not all environments.
Information architecture is clear with README, docs links, contributing guide. Install notes provide quick start steps. Naming stability is not explicit. Examples and FAQ are sparse. Known limitations are not mentioned. License is MIT, but versioning/changelog is not detailed. Maintenance responsibility is not clear.
Output usability is addressed with API and SDKs, but actual output not verified. Marginal value is high with rich features, but cost-benefit not assessed.
Claim traceability is weak as README claims not directly mapped to code. Cross-source corroboration is insufficient. Fact-inference separation is not explicit.
- Not found in source: least-privilege scopingGrant only what the task needs: a dedicated account or read-only token, scoped to specific directories and repos.
- Not found in source: confirmation before actingTurn on (or add) a confirmation step before it acts, and try it in a sandbox or test environment before real data.
- Not found in source: data-flow disclosureWatch which external services it contacts (proxy or firewall logs) and keep sensitive data out until you know where it goes.
- Not found in source: sensitive-data handlingUse dedicated, low-privilege, revocable API keys — never production credentials — and keep secrets out of logs.
- Not found in source: dependency securityPin versions and run a dependency audit (npm audit, pip-audit) before installing; prefer running it in a container.
- Not found in source: disclosed external effectsEstablish which external systems it writes to, sends to or changes, and verify with test accounts or repos before production.
- Not found in source: rollback or recovery pathBack up first, or work on a git branch or snapshot, so its changes can be undone.
- Not found in source: verifiable attributionInstall from the official repo or registry and check the publisher and URL to avoid look-alike packages.
- No evidence of permission management or data security measures; self-hosting requires own assessment.
- Dependency security not audited; recommend security scan before use.
- Known limitations not documented; potential issues may be hidden.