BYOK AI router & FinOps control plane: audit token waste, check tool safety, and rank models.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Requests go through mcptry because browsers can't call most MCP servers directly. Header values are sent to the server for that request only and are never stored or logged.
Tool list from our last check. Connect live to refresh it.
Analyze past LLM request logs (raw JSON/JSONL/CSV or array of message turns) to detect cost waste, unnecessary context bloat, prompt-cache opportunities, and achievable savings with SentioGate.
Array of past chat completion request/response records with token counts and model names.
Raw string containing JSON lines, CSV, or formatted request history logs.
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http sentiogate-mcp https://api.sentiogate.dev/v1/mcp
Analyze past LLM request logs (raw JSON/JSONL/CSV or array of message turns) to detect cost waste, unnecessary context bloat, prompt-cache opportunities, and achievable savings with SentioGate.
| Argument | Type | Description |
|---|---|---|
| records | array | Array of past chat completion request/response records with token counts and model names. |
| raw_logs | string | Raw string containing JSON lines, CSV, or formatted request history logs. |
{
"type": "object",
"properties": {
"records": {
"type": "array",
"items": {
"type": "object"
},
"description": "Array of past chat completion request/response records with token counts and model names."
},
"raw_logs": {
"type": "string",
"description": "Raw string containing JSON lines, CSV, or formatted request history logs."
}
}
}Route a chat completion sub-task through SentioGate (using model "auto" or a pinned provider/model) or enqueue it on the background batch queue.
| Argument | Type | Description |
|---|---|---|
| model | string | Model ID (default: "auto" for automatic cost/tier routing). |
| prompt* | string | The user prompt or sub-task instruction to execute. |
| system | string | Optional system instruction. |
| async_batch | boolean | When true, enqueues the request on the async batch queue (/v1/batch/enqueue) instead of waiting synchronously. |
{
"type": "object",
"required": [
"prompt"
],
"properties": {
"model": {
"type": "string",
"description": "Model ID (default: \"auto\" for automatic cost/tier routing)."
},
"prompt": {
"type": "string",
"description": "The user prompt or sub-task instruction to execute."
},
"system": {
"type": "string",
"description": "Optional system instruction."
},
"async_batch": {
"type": "boolean",
"description": "When true, enqueues the request on the async batch queue (/v1/batch/enqueue) instead of waiting synchronously."
}
}
}Fetch real-time SentioGate BYOK FinOps metrics: measured spend by provider, prompt-cache USD savings, active sessions, and remaining 5-hour Community Pool quota.
| Argument | Type | Description |
|---|---|---|
| time_range | string | Time window for cost analytics aggregation (default: 7d). |
{
"type": "object",
"properties": {
"time_range": {
"enum": [
"today",
"7d",
"30d"
],
"type": "string",
"description": "Time window for cost analytics aggregation (default: 7d)."
}
}
}Pre-flight check tool calls, terminal commands, or conversation history against SentioGate Armor security rules (credential leakage, destructive rm, pipe-to-shell, git force-push, and runaway tool loops).
| Argument | Type | Description |
|---|---|---|
| messages | array | Optional conversation messages array to check for stuck/repeating agent tool loops. |
| arguments* | string | JSON-encoded arguments or raw shell command string. |
| tool_name* | string | Name of the tool (e.g. bash, exec_command, curl, send_email). |
{
"type": "object",
"required": [
"tool_name",
"arguments"
],
"properties": {
"messages": {
"type": "array",
"items": {
"type": "object"
},
"description": "Optional conversation messages array to check for stuck/repeating agent tool loops."
},
"arguments": {
"type": "string",
"description": "JSON-encoded arguments or raw shell command string."
},
"tool_name": {
"type": "string",
"description": "Name of the tool (e.g. bash, exec_command, curl, send_email)."
}
}
}Preview and simulate context compaction and tool-output trimming on an array of conversation messages without executing upstream model inference.
| Argument | Type | Description |
|---|---|---|
| messages* | array | Array of OpenAI-format chat messages to test optimization on. |
{
"type": "object",
"required": [
"messages"
],
"properties": {
"messages": {
"type": "array",
"items": {
"type": "object"
},
"description": "Array of OpenAI-format chat messages to test optimization on."
}
}
}Recommend the optimal model and provider fallback chain for a given coding or agent task based on real-time task difficulty, live catalog pricing, and measured latency.
| Argument | Type | Description |
|---|---|---|
| priority | string | Routing optimization objective (default: cost). |
| task_prompt* | string | The user prompt or instruction to classify and score. |
| requires_tools | boolean | Whether the agent workload requires native tool/function calling. |
{
"type": "object",
"required": [
"task_prompt"
],
"properties": {
"priority": {
"enum": [
"cost",
"latency",
"quality"
],
"type": "string",
"description": "Routing optimization objective (default: cost)."
},
"task_prompt": {
"type": "string",
"description": "The user prompt or instruction to classify and score."
},
"requires_tools": {
"type": "boolean",
"description": "Whether the agent workload requires native tool/function calling."
}
}
}We read every tool's name, description and inputs the way the model will, looking for instructions hidden from you. A clean result means no warning signs, not a guarantee. Scanned Nov 8, 2026.