Watchlist tracking
Reviewed for relevance and future tracking value.
MCP100 Assessment
Watchlist / Tracking
Why we’re tracking it
Follow package releases and MCP tool abstractions as the SDK changes
Reviewed · Documentation review; complete setup workflow not independently tested
Best fit
TypeScript teams building durable multi-provider agents and swarms
What we know
Zero-dependency TypeScript SDK for models, agents, swarms, durable execution, memory, hybrid RAG, MCP tool calling, approvals and CodeAct sandboxes.
Maintenance / activity evidence
- Observed (UTC)chore(bench): remove the 1.8.0 research notes
Tracking evidence observed Oct 5, 2026. These dated signals do not establish current maintenance. Reviewed enrichment confidence: high.
Verification gaps / watch-outs
- Installation and comprehensive security have not been independently audited.
- Application-defined tool and sandbox authority. Writes: Depends on configured MCP/tools/sandboxes.
- Provider/tool-specific.
- Sensitive access: Provider credentials, Application data
- The SDK can orchestrate tools, durable actions and sandboxes; application policy and approvals define the boundary.
Full authority, access and evidence limits are available in Review context and Access & risk.
Popularity
697 repository stars
Stars observed Sep 25, 2026.
Preserve any already-reviewed adoption metrics in MCP100. Batch 4 did not refresh package downloads because the package-stats provider had insufficient credits.
Preserve any already-reviewed adoption metrics in MCP100. Batch 4 did not refresh package downloads because the package-stats provider had insufficient credits.
All limitations / Review context
Which MCP tool interfaces and agent runtimes does the current SDK support?
The bounded assessment reviewed the official README, repository notices, security policy path, and security page for known material public-tracking blockers. No such blocker was identified. This is a public-tracking assessment, not an installation safety certification.
- The SDK can orchestrate tools, durable actions and sandboxes; application policy and approvals define the boundary.
- Application-defined tool and sandbox authority.
- Provider/tool-specific.
- Prompts/data flow to configured model providers and external tools.
- Reviewed write actions: Depends on configured MCP/tools/sandboxes
- Sensitive access: Provider credentials, Application data
- Installation, client compatibility and comprehensive security have not been independently audited. Authentication and access depend on the configured services.
- A project-specific SECURITY.md policy was not available at the fetched official path; the review remains limited to published repository notices.
- No installation, runtime, client matrix, or complete tool-contract audit was performed in this review.
- The fetched official SECURITY.md path returned HTTP 404; it was not treated as positive evidence.
- Provider/tool/sandbox risk depends on application configuration.
Original activity observations
- 2026-09-26T18:59:19Z: chore(bench): remove the 1.8.0 research notes
Technical identity
github.com/deuz-ai/deuz-sdk- Exact implementation source
- https://github.com/deuz-ai/deuz-sdk ↗
- Reviewed resource type
- SDK
- Repository identity
- deuz-ai/deuz-sdk
- Package identity
- npm:@deuz-sdk/core
Evidence & Sources
- Evidence basis
- Recorded source evidence
- Setup runtime verification
- Not independently runtime-tested by MCP100 for this setup workflow.
Reviewed · Latest source observation
- GitHub repository ↗
deuz-ai/deuz-sdk
Recorded repository material and project history.
Includes references whose attribution is not established.
Setup · Watchlist
SDK setup
Publisher instructions.
Source evidence observed .
Additional setup details required — see the specific gaps below.
Install / connect
Publisher instructionsPublisher-specific setup path. Follow prerequisites, then register or connect to your MCP host.
publisher documented install or connect · Publisher instructions · Review before use
npm install @deuz-sdk/coreConfigure / authenticate
Publisher instructionsDeuz is a TypeScript AGENT SDK supporting MCP tool calling, NOT a standalone MCP stdio server. Import and configure @deuz-sdk/core inside a Node/Bun app, with provider credential as required.
Check connection
Suggested checkRun your application integration test and verify configured MCP tools through the SDK; installing the npm library by itself is not an MCP connection.
Remove / uninstall / revoke
MCP100 setup guideRemove the @deuz-sdk/core dependency if the app no longer uses it and delete its MCP tool configuration in your own application; no MCP host `remove` command applies to the SDK.
What it does
Zero-dependency TypeScript SDK for models, agents, swarms, durable execution, memory, hybrid RAG, MCP tool calling, approvals and CodeAct sandboxes.
- Capability
- AI & Reasoning
- Resource type
- SDK
- Industry
- Software & Technology
- Best fit
- TypeScript teams building durable multi-provider agents and swarms
Representative uses
- Build provider-agnostic agents
- Use MCP tool calling
- Add memory/RAG
- Require human approval
- Run CodeAct sandboxes
Limitations
- Provider/tool/sandbox risk depends on application configuration.
Reviewed access
Access & risk
Review the exact access boundary before connecting this MCP to accounts, credentials, or sensitive data.
- Authentication
- Provider/tool-specific.
- Permissions
- Application-defined tool and sandbox authority.
- Data handling
- Prompts/data flow to configured model providers and external tools.
- Cost
- usage_based
- Risk and cautions
- The SDK can orchestrate tools, durable actions and sandboxes; application policy and approvals define the boundary.
Access scope and cautions
- Write actions
- Depends on configured MCP/tools/sandboxes
- Sensitive access
- Provider credentials Application data
- Risk level
- high
FAQ
Based on reviewed evidence as of 2026-10-06. Watchlist tracking is not a recommendation or a security certification.
Which tasks are documented for this implementation?
What authentication is documented?
Can it modify data or perform write operations?
What data-handling boundary should I consider?
What limitations and risks should I check before use?
What cost model is recorded?
Is this an official MCP?
Deuz AI is listed as First-party MCP in MCP100’s current evidence.