Watchlist tracking
Reviewed for relevance and future tracking value.
MCP100 Assessment
Watchlist / Tracking
Why we’re tracking it
Follow Ollama and transport compatibility as local model clients change
Reviewed · Documentation review; complete setup workflow not independently tested
Best fit
Running local Ollama or OpenAI-compatible models with multiple MCP servers
What we know
Terminal MCP client/harness for Ollama and other providers with tools, prompts, resources, multi-server support and human-in-the-loop controls.
Maintenance / activity evidence
- Observed (UTC)fix(github): repair issue forms and improve the PR template ( 329)
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.
- Inherits capabilities of connected MCP servers. Writes: Depends on connected MCP tools.
- Provider-specific for cloud models.
- Sensitive access: Prompts, MCP results, Provider credentials
- The host's practical risk comes from connected MCP servers and selected model providers.
Full authority, access and evidence limits are available in Review context and Access & risk.
Popularity
826 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 transports and Ollama versions does the current client 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 host's practical risk comes from connected MCP servers and selected model providers.
- Inherits capabilities of connected MCP servers.
- Provider-specific for cloud models.
- Local with Ollama; cloud-provider mode sends prompts/tool context to the selected provider.
- Reviewed write actions: Depends on connected MCP tools
- Sensitive access: Prompts, MCP results, Provider credentials
- 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.
- Connected MCP servers are trusted by the user while their outputs are treated as untrusted model input.
- Cloud providers require separate keys and external data flow.
Original activity observations
- 2026-10-01T14:53:06Z: fix(github): repair issue forms and improve the PR template ( 329)
Technical identity
github.com/jonigl/mcp-client-for-ollama- Exact implementation source
- https://github.com/jonigl/mcp-client-for-ollama ↗
- Reviewed resource type
- Host
- Repository identity
- jonigl/mcp-client-for-ollama
- Package identity
- pypi:ollmcp
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 ↗
jonigl/mcp-client-for-ollama
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 instructionsThis is **an MCP client**, not an MCP server. Publisher: `uv tool install --upgrade ollmcp`, then `ollmcp mcp add playwright -- npx @playwright/mcp@latest` to register a *different* MCP server. Run Ollama separately.
package install · Publisher instructions
uv tool install --upgrade ollmcpConfigure application / integrations
Publisher instructionsollmcp is an MCP client/TUI, not a server to register. Install with uv tool install --upgrade ollmcp; requires local Ollama, Python 3.11+ and uv; manage downstream servers with ollmcp mcp add.
Check connection
Suggested checkRun ollama list to verify the backend, inspect ollmcp --help; do not invent a server launch.
Disconnect / uninstall / revoke
MCP100 setup guideFirst use `ollmcp mcp list` then `ollmcp mcp remove <server-name>` to unregister the separate upstream tools (confirm this command against installed version). Removing ollmcp itself does not uninstall Ollama.
Distinguish client registration removal from software uninstall, persistent data deletion and provider OAuth/API-key revocation.
uv tool uninstall ollmcp
What it does
Terminal MCP client/harness for Ollama and other providers with tools, prompts, resources, multi-server support and human-in-the-loop controls.
- Capability
- AI & Reasoning
- Resource type
- Host
- Industry
- Software & Technology
- Best fit
- Running local Ollama or OpenAI-compatible models with multiple MCP servers
Representative uses
- Connect Ollama to MCP servers
- Run agent mode
- Hot-reload multiple servers
- Switch among local/cloud model providers
Limitations
- Connected MCP servers are trusted by the user while their outputs are treated as untrusted model input.
- Cloud providers require separate keys and external data flow.
Reviewed access
Access & risk
Review the exact access boundary before connecting this MCP to accounts, credentials, or sensitive data.
- Authentication
- Provider-specific for cloud models.
- Permissions
- Inherits capabilities of connected MCP servers.
- Data handling
- Local with Ollama; cloud-provider mode sends prompts/tool context to the selected provider.
- Cost
- usage_based
- Risk and cautions
- The host's practical risk comes from connected MCP servers and selected model providers.
Access scope and cautions
- Write actions
- Depends on connected MCP tools
- Sensitive access
- Prompts MCP results Provider credentials
- Risk level
- moderate
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?
jonigl is listed as Community implementation in MCP100’s current evidence.