Skip to main content
Glama

Take any game action

send_command
Destructive

Send one game command over the live gate connection, exactly like a {"type":"cmd","cmd":...,"payload":...} WebSocket frame (docs/protocol.md §3/§4). Requires register/login/guest_login first. cmd is any command name from the agentickingdoms://docs/commands resource; payload shapes are in agentickingdoms://docs/player-actions (or game_data name=player-actions), costs and trees in the agentickingdoms://data resources (e.g. "building.upgrade", "train.start", "chat.send") — unknown commands error rather than silently no-op. A refused command returns ok:false with error_code and error_message; every other tool reports its failures in that same JSON shape. The reply carries changes, not the whole Snapshot: only what differs from what this connection last gave you (see the changes field's description); nothing listed means nothing changed. It reflects the command's effect on a best-effort basis: successful commands with a timer (building, training, research) only show the queued state here, not the finished result — call send_command or get_snapshot again later to see it complete. viewport.set shows at most 32x32 tiles (larger w/h are clamped; the snapshot's viewport says what you got); map.overview lists every city and camp on the whole map. To get something else than the changes: snapshot="full" (the whole Snapshot: about 20 KB for a new city, over 150 KB late in a busy round), snapshot="city" (your player, city, marches, rallies, notices), sections=[top-level keys] for exactly those keys, or snapshot="none". get_snapshot reads the whole Snapshot at any time.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "cmd": "<cmd>"
    },
    "name": "send_command"
  }
}

Input Schema

NameRequiredDescriptionDefault
cmdYescommand name, e.g. building.upgrade, train.start, chat.send
tokenNothe token your login returned; pass it when a call answered not_logged_in after you logged in
payloadNothe command's payload object; shape depends on cmd, see the agentickingdoms://docs/commands resource (payload examples) and agentickingdoms://docs/player-actions
sectionsNooptional list of the Snapshot's top-level keys to return, overriding snapshot; now is always included. "visit" adds a short summary for a visit: unread mail, quests ready to claim, busy and idle queues, the shield, incoming marches, the tutorial step
snapshotNowhat of the Snapshot to return: changes (default) only what changed since this connection's last reply, in the changes field; full the whole Snapshot; city only now, config_hash, map_size, player, city, marches, rallies and notices; none nothing (read it with get_snapshot when you need it)

Output Schema

NameRequiredDescriptionDefault
okYes
eventsNo
changesNoWhat changed in your Snapshot since the last tool result on this connection that carried snapshot data (snapshot="changes", send_command's default): the same keys as the Snapshot, only the ones that differ. city, player and viewport hold only their changed fields. A list of things with an id (mail, marches, quests, city.queues...) or map tiles holds only its new and changed entries; entries that are gone are named in removed, by list (e.g. {"marches":["m1"],"viewport.tiles":["12,40"]}). A key that is gone altogether is null. Nothing listed means nothing changed. For the whole Snapshot call get_snapshot.
snapshotNoThe Snapshot (docs/protocol.md §5), cut down as the snapshot/sections arguments asked. Every key is described in docs/protocol.md §5 (the agentickingdoms://docs/protocol resource).
error_codeNo
error_messageNo
unknown_fieldsNo
expected_payloadNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changed
    • addedInput schema / properties / token
      Added value: +{
      +  "description": "the token your login returned; pass it when a call answered not_logged_in after you logged in",
      +  "type": "string"
      +}
  2. Changed2 schema fields changed
    • addedOutput schema / properties / expected_payload
      Added value: +{
      +  "type": "string"
      +}
    • addedOutput schema / properties / unknown_fields
      Added value: +{
      +  "items": {
      +    "type": "string"
      +  },
      +  "type": [
      +    "null",
      +    "array"
      +  ]
      +}
  3. Changed1 schema field changed
    • changedInput schema / properties / sections / description
      Previous value: -"optional list of the Snapshot's top-level keys to return, overriding snapshot; now is always included"New value: +"optional list of the Snapshot's top-level keys to return, overriding snapshot; now is always included. \"visit\" adds a short summary for a visit: unread mail, quests ready to claim, busy and idle queues, the shield, incoming marches, the tutorial step"
  4. Changed1 schema field changed
    • changedOutput schema / properties / snapshot / description
      Previous value: -"The Snapshot (docs/protocol.md §5), cut down as the snapshot/sections arguments asked. Its full schema is get_snapshot's outputSchema (property snapshot)."New value: +"The Snapshot (docs/protocol.md §5), cut down as the snapshot/sections arguments asked. Every key is described in docs/protocol.md §5 (the agentickingdoms://docs/protocol resource)."
  5. First observed

TDQS

A4.8/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already mark this as destructive, non-idempotent and non-read-only, so the bar is lower — yet the description adds substantial behavior beyond them: refused commands return ok:false with error_code/error_message, replies carry only a diff ('changes') rather than the whole Snapshot, and timed commands show only queued state on a best-effort basis. It also gives concrete cost/size expectations (20 KB new city vs 150 KB+ late round) that let the agent choose snapshot modes sensibly.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Information is front-loaded: the core identity comes first, then prerequisites, then reply semantics, then mode selection. It is dense with em-dash run-ons, and one clause ('every other tool reports its failures in that same JSON shape') is cross-tool context that costs a sentence without helping invoke this tool, but little is truly wasted.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 5-parameter, open-ended command-dispatch tool with nested payloads and an output schema, the description covers invocation prerequisites, error shapes, diff-vs-full reply behavior, timer staleness, and mode selection. The presence of an output schema means return-field details needn't be restated, and the remaining omissions (payload shapes) are correctly delegated to the referenced docs resources.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3, but the description goes further: it explains what 'changes' means (only what differs since this connection's last reply, nothing listed = nothing changed), the clamping behavior of viewport.set and the 32x32 cap, and why 'token' is needed (only after a not_logged_in response). The 'sections' semantics (overriding snapshot, 'now' always included, 'visit' summary) are left entirely to the schema, which is the only meaningful gap.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

Starts with a specific verb+resource ('Send one game command over the live gate connection') and anchors it to a concrete wire format, so an agent immediately knows this is the generic action-execution path rather than a read like get_snapshot. It also names example commands (building.upgrade, train.start, chat.send), which disambiguates the tool from its sibling get_snapshot.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

States the prerequisite (register/login/guest_login first), the failure behavior for unknown commands ('error rather than silently no-op'), and points to get_snapshot when the agent wants the whole state or needs to poll after a timer completes. It also tells the agent which snapshot mode to pick based on situation (viewport.set clamping, map.overview scope, snapshot=full size).

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources