@battlegrid/mcp-server 31.2.48 → 31.2.50

package/README.md

@@ -24,6 +24,110 @@ Seeing package `31.x` alongside handshake `battlegrid@33.x` — the package **be
 
 **What this changes for you:** nothing about how you call anything. Upgrading the package no longer waits on a server deploy, and a server deploy no longer strands you on a package that names the wrong contract — reconnect and the announcement follows. **Contract breaking-change notes are no longer keyed to package versions**, since a contract move is no longer a release here; the v11-and-earlier notes below are kept as history, and the live vocabulary is always discovery.
 
+## Contract history — v68.1 (Arena regime sets)
+
+Purely additive in schema, with **one refusal behind unchanged ones.** An Arena rule's regime condition
+is a SET in the draft, as it always was in the committed deployment: one rule can fire in several
+regimes, instead of one rule per regime.
+
+### Wider input — `stage_deployment_policy_draft`
+
+- **A rule's `regimes` may name several regimes**, where the draft capped it at one. The set is stored in
+  the selectable-regime order, so a staged `[VOLATILE, BULL_EXPANSION]` reads back
+  `[BULL_EXPANSION, VOLATILE]`. A regime named twice is refused.
+
+### Wider output — `get_deployment_policy_draft`
+
+- **The draft may return such a set.** `get_deployment_policy` has served sets all along, so a client that
+  reads a committed deployment already handles them.
+
+### Refusals behind unchanged schemas
+
+- **`preview_deployment_resolution` (SLOTS), `preview_radar_resolution` (SLOTS) and
+  `test_generate_deployment_grid` refuse a regime condition naming one regime twice**, as
+  `VALIDATION_ERROR` at its `regimes` field, where they used to resolve or generate over it. The stage tools
+  already refused a repeat, so no commit changes; a repeated member is a malformed set.
+
+## Contract history — v68 (Arena drafts)
+
+**Breaking: one tool retired, two inputs reshaped.** Arena deployment content now commits only as a
+draft the player was shown a live preview of, as radar content has since v66. The player's deploy
+editor and every conversation share one unsaved draft per arena.
+
+### Removed tool — something you call no longer exists
+
+- **`upsert_deployment_policy` is retired**, with no alias. Calling it is an unknown-tool error. To
+  deploy or change an arena's deployment:
+  1. `stage_deployment_policy_draft` the axes you are changing (`RULES`, `DEFAULT_SLOT`,
+     `REGIME_ANCHOR`), with the `draftVersion` your last read returned;
+  2. `preview_deployment_resolution` with `request: { kind: "DRAFT", draftVersion }` at the version
+     staging returned;
+  3. show the player the resolution and whether it will play (`enabledAfterCommit`), and on their word
+     call `commit_deployment_policy_draft` with the `previewToken` that preview returned and
+     `confirm: true`.
+
+  Pausing is no longer a flag on the write: use `pause_deployment_policy`. **A first deployment always
+  plays**; a replacement keeps its pause.
+
+### Rejected input — something you send is no longer accepted
+
+- **`preview_deployment_resolution`'s `request` needs a `kind`.** `{ kind: "SLOTS", slots, … }` is the
+  old request. `{ kind: "DRAFT", draftVersion }` previews the player's draft of the arena at that
+  version, composed over the deployed policy exactly as the commit writes it. `{ kind: "COMMITTED" }`
+  previews the deployed policy a resume would arm. A request without `kind` is refused.
+- **`delete_deployment_policy` requires `expectedPolicyId`** beside `expectedRevision`: a revision
+  restarts at 1 when an arena is redeployed, so only the pair names the deployment you read.
+- **A regime anchor override names its coin and timeframe together, or neither**, on every path
+  including `test_generate_deployment_grid` and the SLOTS preview.
+
+### Wider surface — eight tools added
+
+- **`stage_deployment_policy_draft`**, **`get_deployment_policy_draft`**,
+  **`list_deployment_policy_drafts`** and **`discard_deployment_policy_draft`** (`mcp:read`). None of
+  them commits or arms anything.
+- **`commit_deployment_policy_draft`** (`presetId`, `previewToken`, `confirm: true`; `mcp:wager`) — the
+  only MCP committer of Arena content.
+- **`pause_deployment_policy`** (`presetId`; `mcp:wager`) — no revision and no certificate.
+- **`resume_deployment_policy`** (`presetId`, `previewToken`, `confirm: true`; `mcp:wager`) — needs the
+  certificate a live `COMMITTED` preview returned.
+- **`cancel_market_grid_submission`** (`sessionId`, `confirm: true`; `mcp:wager`) — the player's own
+  cancellation and refund of one entry.
+
+### Reshaped and wider output
+
+- **`preview_deployment_resolution`** nests the resolution under `resolution` and gains
+  **`previewToken`** (the certificate a live DRAFT or COMMITTED preview earns, null otherwise, good for
+  five minutes) and **`enabledAfterCommit`**.
+- **`pause_deployment_policy` and `delete_deployment_policy` return `openEntries`**: the entries the
+  player's agent already made in the arena's pending sessions, each with its lock time and entry fee.
+  They play out unless the player cancels them — tell the player and ask, then call
+  `cancel_market_grid_submission` only for the entries they pick.
+- **`get_deployment_policy`'s `authoringContext`** gains the arena's header facts
+  (`regimeReferenceTicker`, `presetBadgeImageUrl`, `entryFee`, and `playerCount` — null when the arena
+  has no pending session to count).
+- The thought-log **`outcome`** gains **`SKIPPED_DEPLOYMENT_DISARMED`**: an entry job that found its
+  deployment withdrawn, paused or no longer slotting its agent stood down before paying.
+
+### Refusals worth knowing before you commit or resume
+
+- **The certificate is bound to what you previewed**: your credential, the player, the arena, the
+  subject, the deployment and its revision, the draft version and the composed content. Anything that
+  moves is a `CONFLICT` and nothing is written. Preview again; never reuse the old certificate.
+- **A commit never withdraws**: a draft with no rule and no catch-all is refused, naming
+  `delete_deployment_policy`.
+- **Resume is refused on a retired arena and on one the player's access was revoked from**; pausing and
+  withdrawing stay open.
+- **Deployment refusals answer `INVALID_DEPLOYMENT_POLICY`.**
+- **`delete_deployment_policy` also ends the player's draft of the arena.**
+
+### Vocabulary
+
+`toolCount` goes 134 → 141 (eight added, one retired). `preview_deployment_resolution`'s input and
+output, `delete_deployment_policy`'s input and output, `get_deployment_policy`'s and
+`list_deployment_policies`' outputs, the three journal outputs and `test_generate_deployment_grid`'s
+input (a session start now accepts up to 288 times a day) move; `upsert_deployment_policy`'s are
+removed.
+
 ## Contract history — v66.1 → v67
 
 **Breaking at v67: four output fields removed.** v66.1 and v66.2 were additive and are recorded here

package/dist/index.d.ts

@@ -49,7 +49,7 @@ import { type Implementation, type Prompt, type Resource } from '@modelcontextpr
  * being asked. Move it for a change to THIS package — a proxy fix, a dependency bump, a docs
  * correction. Never move it to track the server.
  */
-export declare const PACKAGE_VERSION = "31.2.48";
+export declare const PACKAGE_VERSION = "31.2.50";
 export declare const DEFAULT_URL = "https://mcp.battlegrid.trade/mcp";
 export interface EnvConfig {
     apiKeys: string[];

package/dist/index.js

@@ -52,7 +52,7 @@ import { ListToolsRequestSchema, CallToolRequestSchema, ListPromptsRequestSchema
  * being asked. Move it for a change to THIS package — a proxy fix, a dependency bump, a docs
  * correction. Never move it to track the server.
  */
-export const PACKAGE_VERSION = '31.2.48';
+export const PACKAGE_VERSION = '31.2.50';
 export const DEFAULT_URL = 'https://mcp.battlegrid.trade/mcp';
 const MAX_RETRIES = 3;
 const RETRY_DELAYS_MS = [2000, 4000, 8000];

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@battlegrid/mcp-server",
-  "version": "31.2.48",
+  "version": "31.2.50",
   "description": "BattleGrid MCP server — play crypto prediction games from AI agents",
   "type": "module",
   "main": "dist/index.js",

package/skills/EXPORT.json

@@ -1,11 +1,11 @@
 {
   "generator": "battlegrid-app/server/scripts/export-mcp-skills.mjs",
-  "contractVersion": "68.0.0",
+  "contractVersion": "68.1.0",
   "files": {
     "battlegrid-agent-management/SKILL.md": "fbc602e9445d0f72353bfba23c30a364b478bd04a9cc6248418cfeec66717eb1",
     "battlegrid-arena-play/SKILL.md": "4cfaa92a8d0dcaf477b227abd774e7aa8642c185d72108dfd401be0cf8aecd7f",
     "battlegrid-market-analysis/SKILL.md": "22fde9c4eac0261c89a8056ab8b20b66fcbe1fd58447fa0ac84fd410b4d134d7",
-    "battlegrid-radar-deployment/SKILL.md": "e2098ac23d80ab985eb93d46f6cc81c84b63dbba12b3bb3eb4cfdcef10ef3b82",
+    "battlegrid-radar-deployment/SKILL.md": "87601043eb3920faa153061e4ea93f4d8618dbdb8fe8c3be52b4e47ddba18d83",
     "battlegrid-strategy-authoring/SKILL.md": "95e91dd1ff36922b7ed0f5eb92c98def98fa39cce37c3d7447269aea1b5a017f",
     "battlegrid-strategy-doctor/SKILL.md": "d0b9e9e07648246d5a23b0c17183278b70b5bcbb5f472ac7da13d03e42a6c6c6",
     "battlegrid-strategy-examples/SKILL.md": "17e8cf8040813622bd5b82165c7af82e03287fee1f6aac974e0b6c584580dd27",

package/skills/battlegrid-radar-deployment/SKILL.md

@@ -84,7 +84,7 @@ it certifies nothing and cannot be committed.
 
 1. `stage_deployment_policy_draft` with `draftVersion` = the `version` you read (0 when there was no
    draft) and only the axes you are changing, each WHOLE: `RULES` (the complete ordered rule list —
-   first is highest priority, one regime per rule), `DEFAULT_SLOT` (the catch-all, or null for none)
+   first is highest priority; a rule's regime condition names a set), `DEFAULT_SLOT` (the catch-all, or null for none)
    and `REGIME_ANCHOR` (the anchor override, or null to inherit the arena's). Pausing is never staged.
 2. `preview_deployment_resolution` with `request: { kind: "DRAFT", draftVersion }` at the version
    staging returned.