# BoxLang AI (slim)

> BoxLang AI (bx-ai) is Ortus Solutions' Apache-2.0 agent framework for BoxLang on the JVM, with agents, tools, memory, RAG, human approval, and an MCP client and server across more than 15 model providers.

- Full: https://www.anchorterminal.com/tools/boxlang-ai.md (~8,150 tokens) · this version ~1,530 tokens · JSON https://www.anchorterminal.com/tools/boxlang-ai.json · canonical https://www.anchorterminal.com/tools/boxlang-ai
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · MCP: https://www.anchorterminal.com/mcp · Updated: 2026-11-01 · pricing https://www.anchorterminal.com/tools/boxlang-ai#pricing · reviews https://www.anchorterminal.com/tools/boxlang-ai#reviews
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

**B · 65.8/100 · rank #328 of 1260 · #10 in Agent frameworks & SDKs · not agent-ready · confidence medium**


Assessment: A JVM agent framework with human approval, guardrail middleware and an MCP client over HTTP and stdio, all enforced in code, with three releases in the 90 days to 1 November 2026. It runs only on BoxLang, the stronger guards are opt-in, and minor releases have renamed public classes without a deprecation period.

## Facts

- Kind: Agent framework · vendor: Ortus Solutions · category: Agent frameworks & SDKs · legal entity: Ortus Solutions, Corp. · provenance 74/100
- Packages: forgebox `bx-ai`
- Auth: None · pricing: Free · x402: no · licence: Apache-2.0
- Languages: BoxLang 1.14.0 or later on Java 21 (the 3.6 announcement says BoxLang 1.18 or later)
- Models: OpenAI, Claude, Gemini, Grok, Groq, Mistral, DeepSeek, Cohere, Perplexity, MiniMax, OpenRouter, Hugging Face, AWS Bedrock, Cloudflare Workers AI, Ollama, Docker Model Runner, any OpenAI-compatible API, plus Voyage, ElevenLabs and Cartesia for embeddings and speech
- MCP client: HTTP (JSON-RPC, with event-stream answers read since 3.6.0), and stdio since 3.6.0, with tool filters and prefixes per server
- MCP server: HTTP at /~bxai/mcp.bxm in a web app, and stdio. Basic auth, API keys, CORS and body limits are opt-in
- Multi-agent: Sub-agents and an agent registry
- Durable state: Suspended approval runs checkpoint to cache or JDBC memory and resume without replaying the model call. No crash recovery for runs in general
- Human approval: HumanInTheLoopMiddleware with tool-name, risk-level, annotation, callback and composite policies, through CLI or signed HTTP gateways
- Guardrails: Unicode hygiene and context fencing on by default. Input sanitiser, guardrail, LLM-judge and output-guard middleware opt-in
- Tracing: BoxLang interception points, LoggingMiddleware and FlightRecorderMiddleware. No OpenTelemetry found
- Telemetry: None found in the module's source. Requests go only to the configured providers
- Releases in 90 days: 3.4.0 (1 September), 3.5.0 (11 September) and 3.6.0 (20 October 2026)
- Separate modules: bx-playwright (browser automation with its own MCP server) and bx-audio (audio tools) are Ortus modules announced with 3.6 and aren't graded here
- Paid plans: BoxLang runtime plans, Starter $749 a year, BoxLang+ $1,999 a server or $2,999 a node a year, BoxLang++ $3,999 or $4,999

- Scores: Reliability 66, Performance pending, Schema & documentation 71, Agent ergonomics 74, Security & auth 71, Payments & pricing 50, Task success pending, Maintenance & community 83, Transparency & trust 70 · negative events -3 · total over the 7 assessed categories
- Why: Reliability, Read with the local-software lines. · Schema & documentation, Framework reading. · Agent ergonomics, Framework reading, scored on what a tool-calling agent with MCP takes. · Security & auth, Framework reading. · Payments & pricing, Self-hosted rule, scored on the paid option. · Maintenance & community, 3.6.0 on 2026-10-20 (30). · Transparency & trust, Apache-2.0, OSI approved, though the licence file sits in src/main/resources rather than the repository root (30).
- Sources: 23, open questions: 14, both in the full twin
- Capabilities: agent.framework, agent.multi-agent, agent.mcp-client
- JSON: https://www.anchorterminal.com/api/v1/tools/boxlang-ai.json
- Verify (for the vendor): the badge `https://www.anchorterminal.com/badges/boxlang-ai.svg` or a link to https://www.anchorterminal.com/tools/boxlang-ai from a page on ortussolutions.com or ai.boxlang.io or one of their subdomains, or the README at the root of github.com/ortus-boxlang/bx-ai, then `POST https://www.anchorterminal.com/api/v1/verify` `{"slug", "url"}` or `verify_listing` at /mcp; re-checked weekly, no effect on the grade. Snippets in the full twin.
- Support: https://github.com/ortus-boxlang/bx-ai/issues; https://ai.ortusbooks.com; https://www.ortussolutions.com/services/support; more in the full twin

## Before you call it

1. Set `mcp.stdio.allowedCommands` to the executables you need. The default empty list lets any command start as a stdio MCP server.
2. Pass `allowedPaths` when you create FileSystemTools. With none, its tools can read, write and delete anywhere the JVM user can.
3. Attach HumanInTheLoopMiddleware with a checkpointer for tools that write or spend. Approval is opt-in.
4. Create MCP clients with `MCP()` rather than the class, since 3.6.0 renamed MCPClient to MCPHttpClient without an alias.
5. Use the `mock` provider to test agent wiring and guardrails with no API key or network.

## Connect

```bash
install-bx-module bx-ai   # in a web app: box install bx-ai
```

## Similar tools

| Tool | Grade | Score | Shared capabilities | Slim |
| --- | --- | --- | --- | --- |
| OpenAI Agents SDK | AA | 86.2 | agent.framework, agent.multi-agent, agent.mcp-client | https://www.anchorterminal.com/tools/openai-agents-sdk.min.md |
| Pydantic AI | A | 81.9 | agent.framework, agent.multi-agent, agent.mcp-client | https://www.anchorterminal.com/tools/pydantic-ai.min.md |
| Microsoft Agent Framework | A | 79.4 | agent.framework, agent.multi-agent, agent.mcp-client | https://www.anchorterminal.com/tools/microsoft-agent-framework.min.md |
| Docker Agent | BB | 73.6 | agent.framework, agent.multi-agent, agent.mcp-client | https://www.anchorterminal.com/tools/docker-agent.min.md |
| Agno | BB | 70.9 | agent.framework, agent.multi-agent, agent.mcp-client | https://www.anchorterminal.com/tools/agno.min.md |


## Panel reviews (0, desk reviews from public material, no calls made)