MCP
BattleGrid exposes the same core capability catalogue over MCP that Commander uses in-product — strategies, agents, market reads, Radar, Arena, and trade control from compatible clients outside the UI.
If you already use Commander, MCP is that surface for automation and external agents. Start in the UI; wire MCP when you want the same verbs in your own stack.
How it fits
| You want… | Use |
|---|---|
| Fast plain-language control in the product | Commander |
| Prompt patterns that work well | Prompt formula |
| The same actions from Cursor / other MCP clients | MCP (this page) |
Commander discovers this catalogue at runtime and family-gates it. An MCP client sees the tools your OAuth scopes allow — there is no second, secret API.
Scopes
Two scopes. A family never straddles them.
| Scope | Label (consent) | What it allows |
|---|---|---|
mcp:read | Account & configuration access | Account and market information; create / update / bind / archive / restore non-financial configuration. Cannot submit wagers or move funds. |
mcp:wager | Wager / standing authority | Money movement or standing autonomous authority — Arena/grid submits, Radar deploy upsert/delete (including pause via enabled), agent halt/resume and related runtime wager tools, close position, accept/cancel pending entry decisions, and similar. |
Grant mcp:wager only to clients you trust. Radar upsert is wager-scoped because a policy is standing authority to trade that coin, not a harmless preference.
Destructive tools additionally require an explicit confirm: true (or the tool’s documented confirm field). A missing confirm is a refusal, not a soft no-op.
Capability map (by intent)
Grouped the way Commander skills think — not an exhaustive tool dump.
| Intent | Read side (examples) | Wager / authority side |
|---|---|---|
| Account | Balance, slots, game stats | — |
| Market / coins | Regime, candles, rankings, signal previews | — |
| Strategies | Library, templates, compile/review | Apply plan (configuration; still confirm). Signal-rule updates when bound agents exist need confirm. |
| Agents | List, performance, budget, qualification, activity | Create/update/rebind/archive (config); halt / resume and related runtime (wager) |
| Radar | List/get deployments, preview resolution, activity summary | Upsert / delete deployment (wager) |
| Arena / grids | Presets, policies, preview resolution, generate grid | Upsert/delete deployment policy; submit grid (wager) |
| Live trades | Open positions, entry decisions, charts | Accept/cancel entry; close position; protection overrides (wager) |
Strategy compile parks a plan (not a live mutation). Apply commits with planToken + confirm — five-minute token life, revision-aware. Same contract as Commander. See Strategies.
Authoring contract (strategies)
- Discover vocabulary / section templates as needed.
compile_strategy_planwith a strict CREATE / UPDATE / RESTORE payload → approved plan + review context.- Human (or your agent) reviews blast radius and viability.
apply_strategy_planwith the exactplanTokenandconfirm: true— never rebuild the plan body client-side.
Revision tokens (expectedRevision, sourceRevision, …) are optimistic concurrency. Stale values conflict; nothing is overwritten silently.
Radar and Arena via MCP
- Radar — coin-scoped policy; confirmed regime at deployment timeframe; rule/default slots. Preview is dry-run. Upsert/delete are wager-scoped. Field guide: Radar.
- Arena — game-preset deployment; slots configure game participation only and do not grant trade permission by themselves. To trade a coin, deploy on Radar. Upsert/delete are wager-scoped.
Safety habits
- Prefer read tools until the ask is clear.
- Echo revision / plan tokens byte-for-byte.
- Treat
mcp:wagercalls as irreversible or authority-granting until proven otherwise. - On
mutating-tool-interruptedequivalents (client disconnect mid-write), read back state before retrying — blind retry can double a write. - Use Prompt formula the same way you would in Commander: Intent + Asset + Constraints + Where.
Connecting a client
Authorize an MCP client against BattleGrid with the scopes you intend (mcp:read alone for configuration and research; add mcp:wager only when the client must deploy capital or standing authority). Tool schemas from live discovery are the source of truth for parameter shapes — including the { request: … } envelope on strategy studio tools.