mcp-registry/ai.rokha/rokha · v0.1.0
As of 2026-10-21 14:23 UTC, the Rokha MCP server is up with a Glimind reliability score of 99/100 (grade A). Over the last 30 days it answered 100% of Glimind's liveness checks.
Search a registry of agent skills and MCP servers, then run them for real. No install, traced.
Claude Code
claude mcp add --transport http rokha-mcp-server https://rokha.ai/mcp/jsonrpc
Claude Desktop · Windsurf · Cline (mcpServers JSON)
{
"mcpServers": {
"rokha-mcp-server": {
"type": "http",
"url": "https://rokha.ai/mcp/jsonrpc"
}
}
}
Third-party server, listed as discovered. Review its source before installing.
Loading this server's tool list costs an agent ~67,266 tokens before any work — more than 99% of MCP servers.
Split rarely-used tools into a separate server or a toolset the client can enable, or expose one "search tools" entry point. Many agents load every connected server's tools into every request. How to fix →
1 tool definition(s) exceed ~1,200 tokens each (largest: rig_author, ~4,393).
Move long reference material out of descriptions (link a doc or expose a resource), drop repeated examples, and collapse large enums. Trimming these to ~600 tokens saves ~3,793 tokens every session. How to fix →
291 tools on one server. Tool-selection accuracy drops as the menu grows, and some clients (e.g. Cursor) cap the number of active tools.
Group tools by task into smaller servers or toolsets, and merge near-duplicates behind one tool with a mode parameter. How to fix →
247 parameter(s) across 102 tool(s) have no description.
Add a "description" to each parameter with its format and an example ("ISO 8601 date, e.g. 2026-10-17"). This is the commonest cause of wrong arguments on the first call. How to fix →
4 tool(s) look like they change state (create/send/delete/pay…) but declare no destructiveHint.
Set annotations {"readOnlyHint": false, "destructiveHint": true|false, "idempotentHint": …}. Clients use these to ask the user before a side effect and to auto-approve safe reads. How to fix →
1 tool(s) have parameters with no type (no type/enum/anyOf/$ref).
Give every parameter a "type" (or "enum"/"anyOf"). Untyped parameters make models guess — strings for numbers, objects for arrays. How to fix →
In the last 30 days: 34 tool(s) removed and 105 schema change(s).
Add new tools or optional parameters instead of renaming or removing; deprecate first (say so in the description) and remove later. Agents cache tool contracts. How to fix →
No tool declares an outputSchema.
Declare outputSchema and return structuredContent (MCP 2025-06-18+). Agents can then parse results without guessing, and chain tools reliably. How to fix →
6 tools share an identical description with another tool on this server.
Make each description say how the tool differs from its siblings, or the model will pick between them at random. How to fix →
Runs commands or code supplied by the model. Legitimate for dev tools; clients should require confirmation. Example (sandbox_exec, name): "sandbox_exec(kind, tool, image, action)"
Declare destructiveHint/readOnlyHint on these tools and say plainly in the description what runs, where, and with what limits. How to fix →
Paste-ready definitions with the problems above fixed. Changes marked verify need a human look. 102 tool(s) still need parameter descriptions written by someone who knows the API. All of them as JSON: /v1/readiness/mcp-registry%2Fai.rokha%2Frokha/fixes
{
"name": "task_submit",
"description": "Submit your entry's ONE result before the deadline: {task_id, entry_id, via}. via 'runtime' — Rokha runs the task's capability (or the Market rig you name in `listing`) with the brief as input AS your agent, on your own keys under a per-entry cap; via 'direct' — send {output (≤64 KB), model?} from your own stack; via 'endpoint' — Rokha sends the brief to your agent's registered endpoint. Every result is checked against the output schema, then graded (valid · score · reason). Requires Authorizati",
"inputSchema": {
"type": "object",
"required": [
"task_id",
"entry_id"
],
"properties": {
"via": {
"enum": [
"runtime",
"direct",
"endpoint"
],
"type": "string"
},
"model": {
"type": "string",
"description": "direct: the model that produced it (self-attested)."
},
"output": {
"description": "direct: the result (JSON or text).",
"type": "string"
},
"listing": {
"type": "string",
"description": "runtime, on a task with no capability: the Market rig to run."
},
"task_id": {
"type": "string"
},
"entry_id": {
"type": "string"
}
}
},
"annotations": {
"title": "Submit a result",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": false
},
"title": "Task Submit"
}Get an email or webhook when this server's readiness drops by 10+ points, tool-poisoning text appears, or a tool breaks:
curl -X POST https://glimind.com/v1/subscriptions -H 'content-type: application/json' -d '{"tools":["mcp-registry/ai.rokha/rokha"],"email":"[email protected]"}'Agents: watch_tool({"toolId": "mcp-registry/ai.rokha/rokha"}) over MCP.
rig_author ~4,393 tokenspage_rig_config ~867 tokensrig_node_add ~840 tokenssingularityagent__mesh ~831 tokensaasagenticawesomeskills__export_selection_evidence ~777 tokensStatic checks of the advertised tools/list (schema, descriptions, annotations, tool-poisoning patterns, names, size) plus observed protocol, uptime, latency and contract changes. Read-only: no tool is ever called. Weights: schema 15, docs 20, safety 15, efficiency 15, compatibility 15, protocol 5, reliability 10, stability 5. Token counts are estimates (~3.6 characters per token).
Operator? Fix these, then re-check live. Embed the readiness badge:
[](https://glimind.com/server/mcp-registry%2Fai.rokha%2Frokha?ref=badge)
<a href="https://glimind.com/server/mcp-registry%2Fai.rokha%2Frokha?ref=badge"><img src="https://glimind.com/v1/readiness/badge/mcp-registry%2Fai.rokha%2Frokha.svg" alt="MCP agent-readiness"></a>
.. image:: https://glimind.com/v1/readiness/badge/mcp-registry%2Fai.rokha%2Frokha.svg :target: https://glimind.com/server/mcp-registry%2Fai.rokha%2Frokha?ref=badge :alt: MCP agent-readiness
Glimind records every version of this server's tool definitions in a signed, append-only transparency log. Current manifest sha256:1134d294ecb27992…, first seen 2026-10-19, included at leaf 28876 of 40,003. Version history · verify what your client received.
Compare: most reliable code servers · most reliable data servers · most reliable knowledge servers · most reliable files servers · most reliable search servers
Glimind tells agents whether a server is up right now and how to call it, then fails over to a live alternative when it isn't.
Claude Code
claude mcp add --transport http glimind https://glimind.com/mcp
Claude Desktop · Windsurf · Cline (mcpServers JSON)
{
"mcpServers": {
"glimind": {
"type": "http",
"url": "https://glimind.com/mcp"
}
}
}
As of 2026-10-21 14:23 UTC, the Rokha MCP server is up with a Glimind reliability score of 99/100 (grade A). Over the last 30 days it answered 100% of Glimind's liveness checks.
Rokha MCP server advertises 295 tools, including aasagenticawesomeskills__compose_stack, aasagenticawesomeskills__diff_stack, aasagenticawesomeskills__export_selection_evidence, aasagenticawesomeskills__get_skill, aasagenticawesomeskills__inspect_selection_evidence.
Point your MCP client at this streamable-http endpoint: https://rokha.ai/mcp/jsonrpc
Glimind scanned all 291 tool definitions for tool-poisoning patterns — hidden or invisible characters, instructions that override the model, concealment from the user, requests to read credential or MCP config files, conversation exfiltration and cross-tool shadowing — and found none. This checks what the server tells the model, not what its code does; review the source before granting access to sensitive data.
Status page: 90-day uptime, every outage and how long it lasted.
Claim this listing to get the Verified operator marker, a remediation checklist and owner analytics, or embed the live badge from any of the tool pages above. Maintainers: alerts, deploy attestation and monthly reports.
Status is an automated estimate from Glimind's liveness checks (handshake and tool listing only; no tool is ever called) and may lag or differ from your network. Glimind is not affiliated with this server or its operator. Operator? Request a correction or opt out at https://glimind.com/opt-out.