Watchlist tracking
Reviewed for relevance and future tracking value.
MCP100 Assessment
Watchlist / Tracking
Why we’re tracking it
Track assistant and tool-host compatibility as its local agent loop changes
Reviewed · Documentation review; complete setup workflow not independently tested
Best fit
Self-hosting a personal AI assistant with RAG, browsing, code execution and MCP tools.
What we know
OpenAgent is a self-hostable personal AI assistant that can connect model providers, build document knowledge bases, browse the web, execute code and call MCP tools.
Maintenance / activity evidence
- Observed (UTC)feat: support WeChat mini program sign-in
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.
- Can inherit the capabilities of configured model providers, code execution and MCP tools. Writes: Code/tool actions depend on enabled integrations.
- Deployment-specific.
- Sensitive access: Provider credentials, Knowledge-base documents, Tool-accessible resources
- A self-hosted agent host can aggregate model credentials, local knowledge and powerful tools; deployments need explicit least-privilege configuration.
Full authority, access and evidence limits are available in Review context and Access & risk.
Popularity
5,689 repository stars
Stars observed Oct 6, 2026.
All limitations / Review context
Which MCP client integrations and tool permissions does the current assistant 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.
- A self-hosted agent host can aggregate model credentials, local knowledge and powerful tools; deployments need explicit least-privilege configuration.
- Can inherit the capabilities of configured model providers, code execution and MCP tools.
- Deployment-specific.
- Self-hosted by default, but connected providers/tools may receive data.
- Reviewed write actions: Code/tool actions depend on enabled integrations
- Sensitive access: Provider credentials, Knowledge-base documents, Tool-accessible resources
- Installation, client compatibility and comprehensive security have not been independently audited. Authentication and access depend on the configured services.
- No independent runtime, installation, or comprehensive security audit was performed.
- No installation, runtime, client matrix, or complete tool-contract audit was performed in this review.
- Exact authentication model by deployment
- Provider credentials and MCP tool permissions vary by deployment.
- Code execution and external tools increase the authority of the host.
Original activity observations
- 2026-10-05T17:43:04Z: feat: support WeChat mini program sign-in
Technical identity
github.com/the-open-agent/openagent- Exact implementation source
- https://github.com/the-open-agent/openagent ↗
- Reviewed resource type
- Host
- Repository identity
- the-open-agent/openagent
- Service endpoint
- http://localhost:14000
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 ↗
the-open-agent/openagent
Recorded repository material and project history.
Setup · Watchlist
Host / Desktop app setup
Publisher instructions.
Source evidence observed .
Additional setup details required — see the specific gaps below.
Download / install
Publisher instructionsOpenAgent is an AI assistant application/host.
publisher documented install or connect · Publisher instructions
curl -fsSL https://raw.githubusercontent.com/the-open-agent/openagent/master/scripts/install.sh | bashConfigure application / integrations
MCP100 setup guideRun publisher installer for Linux/macOS or PowerShell installer on Windows; the binary hosts an AI assistant on port 14000. Configure integrations inside application; do not register the app itself as a stdio MCP server.
Check connection
Suggested checkCheck OpenAgent process/service and configured application endpoint before verifying any optional downstream MCP integration.
Disconnect / uninstall / revoke
Suggested checkStop the running OpenAgent executable/service. Remove any MCP integrations separately; delete the installed binary through the installer destination documented by OpenAgent, leaving user data unless explicitly approved.
Client unregistration, program uninstall, persistent-data deletion, and credential revocation must be treated separately.
What it does
OpenAgent is a self-hostable personal AI assistant that can connect model providers, build document knowledge bases, browse the web, execute code and call MCP tools.
- Capability
- AI & Reasoning
- Resource type
- Host
- Best fit
- Self-hosting a personal AI assistant with RAG, browsing, code execution and MCP tools.
Representative uses
- Run a self-hosted assistant
- Build a document knowledge base
- Browse the web
- Connect MCP tools
Limitations
- Provider credentials and MCP tool permissions vary by deployment.
- Code execution and external tools increase the authority of the host.
Reviewed access
Access & risk
Review the exact access boundary before connecting this MCP to accounts, credentials, or sensitive data.
- Authentication
- Deployment-specific.
- Permissions
- Can inherit the capabilities of configured model providers, code execution and MCP tools.
- Data handling
- Self-hosted by default, but connected providers/tools may receive data.
- Cost
- unknown
- Risk and cautions
- A self-hosted agent host can aggregate model credentials, local knowledge and powerful tools; deployments need explicit least-privilege configuration.
Access scope and cautions
- Write actions
- Code/tool actions depend on enabled integrations
- Sensitive access
- Provider credentials Knowledge-base documents Tool-accessible resources
- 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 setup and authentication are 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?
Is this an official MCP?
The Open Agent is listed as First-party MCP in MCP100’s current evidence.