Skip to main content
Glama

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
EMBER_HOSTNoModel server address (host).127.0.0.1
EMBER_PORTNoModel server address (port).8765
EMBER_MODELNoModel to use: flash (9B) or full (27B).flash
EMBER_DEVICENoDevice to run the model on: auto, mps, or cpu.auto
EMBER_AUTOSTARTNoLet the MCP server start the model server on demand.1
EMBER_MODEL_DIRNoRun weights from this directory instead of the pinned cache.
EMBER_STATE_DIRNoWhere the pid file and logs live.Application Support
EMBER_MAX_LENGTHNoToken cap per request; 0 derives it from the model (the model's maximum is 262144).0
EMBER_SERVER_URLNoWhere the MCP server sends requests.http://127.0.0.1:8765
EMBER_START_TIMEOUTNoSeconds to wait for the model server to start.300

Instructions

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

This server publishes no instructions, or was last inspected before Glama recorded them.

Capabilities

Features and capabilities supported by this server

Protocol revision2025-11-25

CapabilityDetails
tools
{
  "listChanged": false
}
prompts
{
  "listChanged": false
}
resources
{
  "subscribe": false,
  "listChanged": false
}

Tools

Functions exposed to the LLM to take actions

NameDescription
adviseA

Get ember's read on a situation, as calibrated probabilities.

ember (Cloudflare's Clef-Flash model, running locally) advises; you decide.
Consult it at bounded decision points: intent, triage, routing, yes/no gates, and
risk, severity, or effort scores. It sees only what you pass, so include every
piece of evidence the call depends on and attach images or video frames as base64
`data:` URIs in `images`/`videos` when pixels are the evidence. For 'choice' the
answer has the leading option, its confidence, and full probabilities; for 'score'
an expected score over the ordered criteria; for 'noul' the probability the
proposition is true.

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription
ember-guideFull playbook for the advise tool: schemas, thresholds, and recipes.

TDQS

A4.4/5.0

Scored across 1 tool

Disambiguation5/5

Only one tool exists, so there is no risk of confusion with other tools; its purpose is clearly stated as getting calibrated probabilistic advice.

Naming Consistency5/5

The single tool name 'advise' is a clear verb, and no inconsistent conventions exist within the set.

Tool Count4/5

One tool is slightly below the typical 3-15 range, but it is well-scoped for a focused advisory endpoint that handles multiple query modes internally.

Completeness5/5

The tool covers choice, score, and noul response types, plus image/video evidence, providing a complete surface for its stated advisory purpose.

Maintenance

ActivityMaintained
ResponsivenessNo issues