@battlegrid/mcp-server 32.0.0 → 32.0.2

package/README.md

@@ -38,6 +38,83 @@ 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 — v118.1 (a server fault says whether repeating is safe)
+
+**Nothing you send is refused that v118 accepted.** v118.1 is the contract on top of v118. No input schema moves.
+Reconnect the proxy, or restart it, so discovery reads the new annotation below.
+
+### Added fields
+
+- **`details.correlationId`** is on every error envelope. The server mints it once per call and writes it on that
+  call's one log entry, so quote it when you report a failure.
+- **`details.retryable`** is on a server fault: `SERVICE_UNAVAILABLE` from an unavailable dependency, and every
+  `INTERNAL_ERROR`. It says whether repeating the same call unchanged is safe, never that it will succeed:
+  - `true` means a repeat has no additional effect. If the repeat is then refused, follow that refusal;
+  - `false` on `SERVICE_UNAVAILABLE` means the call may have taken effect, so read the current state before
+    repeating it;
+  - `false` on `INTERNAL_ERROR` means repeating fails the same way;
+  - absent, on any typed refusal, means its message and `nextAct` say whether repeating helps.
+
+### Reclassified
+
+- **A dependency the server could not use answers `SERVICE_UNAVAILABLE`**, where it answered `INTERNAL_ERROR`. That
+  covers a dropped database or Redis connection, a serialization failure or deadlock, a network timeout, and the
+  exchange budget declining to admit a call. `retryable` is `true` for a tool that declares `readOnlyHint` or
+  `idempotentHint`.
+- **`reset_agent_drawdown_baseline` declares `idempotentHint: false`.** Each call moves the drawdown baseline to now,
+  so a repeat forgives any loss realized in between.
+
+## Contract history — v118 (a refused argument is the house error envelope)
+
+**Breaking at v118: an argument refusal is JSON, and the scope check comes first.** v118 is the contract on top of
+v117. No schema moves, so nothing you send is refused that v117 accepted. Code that read the SDK's refusal sentence
+must read the envelope instead.
+
+### Reshaped output — something you read has a new shape
+
+- **A call whose arguments the tool's schema refuses** is an `isError` result whose text is the same JSON envelope
+  every other refusal uses: `{ "code": "VALIDATION_ERROR", "message", "details": { "issueCount", "issues" } }`. It
+  was the SDK's `Input validation error: …` sentence.
+
+### Refusal vocabulary
+
+- **`details.issues`** holds the first 20 issues in the order the parse found them. Each is `{ path, message }`:
+  - `path` is where the issue is, as segments within the arguments object, and `[]` is the arguments object itself;
+  - `message` is the schema's own text, so a remedy the schema appends to an issue reaches you unchanged.
+- **`details.issueCount`** is how many issues the parse found. It exceeds the length of `issues` when more than 20
+  were found.
+- **An unrecognised key** is reported at the path of the object it was sent on, by an issue whose message names the
+  key. A retired argument therefore shows up as an issue on its parent object, not as a path ending in its own name.
+- **`message`** is the first issue's message.
+- **The refusal carries no `details.nextAct`.** Recover by making the same call with corrected arguments. A refusal of
+  a draft's staged content still carries `nextAct: "stage"`, so the two are told apart by `details.issues`.
+
+### Behaviour behind unchanged schemas
+
+- **The required scope is checked before the arguments are parsed.** A credential without the tool's scope is refused
+  `FORBIDDEN` with `details.requiredScope` whatever it sends. Before, an argument object the schema refused was
+  answered by the parse first.
+
+## Contract history — v117 (decided changes read at their horizon bar)
+
+**Breaking at v117, behind unchanged schemas.** v117 is the contract on top of v116; no input or output schema moves.
+
+### Rejected input — something you send is no longer accepted
+
+- **A verdict carrier or hold that reads a change over a horizon above the strategy timeframe is refused.** That is
+  `chg4h` and `chg24h` on a `1h` strategy, and `oiRegime` (the `1h` change) below `1h`. Such a change is a developing
+  read whatever its selector.
+  - **Where it is refused:** a strategy save, draft stage or preview, refused `CONDITION_VERDICT_READ_ILLEGAL` or
+    `CONDITION_HOLD_ILLEGAL`. Restoring an archived strategy that holds one is refused the same way until it is
+    repaired.
+  - **What the refusal says:** it names the carrier and the column, with a remedy that reads a horizon at or below
+    the strategy timeframe.
+
+### Behaviour behind unchanged schemas
+
+- **A change is the frame's price against its horizon bar's open.** A decided frame reads it exactly where it read
+  UNRESOLVED, and a horizon whose bar was not read is UNRESOLVED where it read FALSE.
+
 ## Contract history — v116 (the server moves to MCP SDK v2 and protocol 2026-07-28)
 
 **Breaking at v116: `test_generate_deployment_grid` takes its payload at the top level.** v116 is the contract on top

package/SKILL.md

@@ -72,8 +72,9 @@ than about authoring:
   `expectedPolicyId` and `expectedRevision` a `COMMITTED` preview returned.
 - **The proxy forwards numbers and `null` unchanged.** It never fills in a version, never retries a
   refused call and never rebuilds a request. A `DRAFT_VERSION_MOVED` or `DRAFT_AXIS_CONTESTED` refusal
-  carries `details.nextAct` from every tool, and a revision `CONFLICT` or validation refusal carries it
-  from a stage or commit; the next act is the caller's.
+  carries `details.nextAct` from every tool, and a revision `CONFLICT` or a refusal of the staged content
+  carries it from a stage or commit. A refused argument (`VALIDATION_ERROR` with `details.issues`)
+  carries none: make the same call with the arguments corrected. The next act is the caller's.
 
 `fork_strategy` copies a revision into a new create draft and creates nothing until that draft is
 committed. An agent binds to a strategy through its draft's `STRATEGY_BINDING` axis, on a create or

package/dist/index.d.ts

@@ -47,7 +47,7 @@ import { Client, StreamableHTTPClientTransport } from '@modelcontextprotocol/cli
  * 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 = "32.0.0";
+export declare const PACKAGE_VERSION = "32.0.2";
 export declare const DEFAULT_URL = "https://mcp.battlegrid.trade/mcp";
 export interface EnvConfig {
     apiKeys: string[];

package/dist/index.js

@@ -49,7 +49,7 @@ import { Client, SdkHttpError, StreamableHTTPClientTransport } from '@modelconte
  * 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 = '32.0.0';
+export const PACKAGE_VERSION = '32.0.2';
 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": "32.0.0",
+  "version": "32.0.2",
   "description": "BattleGrid MCP server — play crypto prediction games from AI agents",
   "type": "module",
   "main": "dist/index.js",

package/skills/EXPORT.json

@@ -1,15 +1,15 @@
 {
   "generator": "battlegrid-app/server/scripts/export-mcp-skills.mjs",
-  "contractVersion": "113.0.0",
+  "contractVersion": "116.0.0",
   "files": {
     "battlegrid-agent-management/SKILL.md": "9dc151c32c3578d06e8895c6752d7a64959bff6a6681a01d62ab65e1ea1590d1",
     "battlegrid-agent-management/references/agent-stage-examples.md": "dffcc353d61dfc7b5bb6a2b3077fc6bf8c760a34c287cafb3b59b99306f24195",
     "battlegrid-arena-play/SKILL.md": "2e37970a66b8738ffbc445a23da90651c1f41b56ce4eef551582ab123517ef74",
     "battlegrid-arena-play/references/entry-stage-examples.md": "0223f3743171ff213e943789d27590b2f65648398e85e10cd24938d5eb6af665",
     "battlegrid-market-analysis/SKILL.md": "fcbe98e30e1edce39bb170c224cdf20181934f62b5f3a26b293787c70eeb6229",
-    "battlegrid-radar-deployment/SKILL.md": "dfe8f354ac3dd426eedeac1ec080ac518e9c8cbfefdd26036afa61f80c08295c",
+    "battlegrid-radar-deployment/SKILL.md": "cf6664f51f92d40acce21a46f4fb078d9afe46c99c4f4b603bf002e73d4c5a23",
     "battlegrid-radar-deployment/references/arena-stage-examples.md": "d27cfed1f4663f267c4a7ecebe4a6e65a450fc74bd14ce8987fb61af5bbf2ca5",
-    "battlegrid-radar-deployment/references/radar-stage-examples.md": "e3805f5c1b243135a7f25fb3a8cc6093dd242f8ad67b7e6b242bbdefa58a0735",
+    "battlegrid-radar-deployment/references/radar-stage-examples.md": "13fc3f55d91d3da9754e1ed7fb27493b7b7ba4d40a6a7e8d1ee0365f90ea10de",
     "battlegrid-strategy-authoring/SKILL.md": "39cfa46b1f640e6800edb7e85a8f2c4b6b8786cddf71bb8244d172a44fc11d75",
     "battlegrid-strategy-authoring/references/strategy-stage-examples.md": "f10ecdc91f3660a708eee7a7bacefce2acbb46c9cf7e0ac1454bd54f839f39fd",
     "battlegrid-strategy-doctor/SKILL.md": "e34e055b0cc9b355bc1adeb5d49d464eaa0756ca3f48dc89ca2ccdda94e1311f",

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

@@ -107,7 +107,12 @@ is test-routing a regime. Never carry a fact from a radar conversation into it,
    Do not confirm a draft whose diagnostics carry an error: stage a fix.
 3. `preview_radar_resolution({ coinId, request: { kind: "DRAFT", draftVersion } })` at the version that
    read returned. It composes the draft over the deployed policy **exactly as the commit will write it**,
-   and refuses a draft no longer at that version with `DRAFT_VERSION_MOVED`.
+   and refuses a draft no longer at that version with `DRAFT_VERSION_MOVED`. Its `ruleCloseCoverage`
+   counts, for each hours rule, the closes of its agent's `strategyTimeframe` bar the rule's hours let
+   it decide through `windowEndAt`. Before asking the player to commit, name every rule whose
+   `decidableCloses` is 0: its agent's timeframe, the served `windowEndAt`, and that it is this rule's
+   hours alone that hold no close — not the agent, which still decides closes it is on duty for through
+   another rule or the default. Never tell the player the agent will not trade.
 
 To try a slot set without touching the draft, preview `{ kind: "SLOTS", slots }` — it resolves
 explicit slots and nothing can commit it.
@@ -122,6 +127,19 @@ from what the player asked for:
 | only in named hours and days | a rule with an hours condition |
 | in named regimes during named hours | a rule with one hours condition and one regime condition |
 
+When the player names an exchange session — New York, the New York open, London, the London open,
+the London–New York overlap, Tokyo, the Tokyo open — call `get_radar_market_hours` and stage that session's `hours`
+exactly as served, merged with any other session they name; never work session hours out yourself,
+because New York and London shift with daylight saving on their own dates and the overlap changes
+shape on each of those dates. The three opens are one hour wide, so they suit agents on 1h bars or
+faster. The days stay the player's choice. Tell the player the staged hours are UTC, that an hour is
+on duty from :00 through :59, and, when the session carries a `nextShift`, the date its hours change
+and that the rule keeps the hours it was given.
+
+Session closes are not presets. Asked for "the New York close" or any other session close, stage only
+the hours the player names, or raise flattening before the session ends as a separate idea — never
+work a close window out yourself.
+
 When the ask maps to no single row, or a refusal rules out the construction the player asked for,
 ask the player which construction they mean, naming the rows that fit — never stage one they did not
 choose. Complete stage payloads for every row are in `references/radar-stage-examples.md`.
@@ -352,8 +370,9 @@ Four reads, cheapest first. Stop as soon as the player's question is answered.
    refused it and nothing was traded. It is the only history of not-qualifying closes: the journal
    writes no row for one, and `resolvesNow` keeps only each agent's last decision — the one
    `reading.lastClose` (step 1) carries; this read holds every decision before it. Narrow with
-   `agentId`; page back by passing the response's `nextBefore` as `before`, and stop when
-   `nextBefore` is null — that is the end of the history. A bar still
+   `agentId`. Each record carries its whole evidence, so keep `limit` small; page back by passing
+   the response's `nextCursor` as `cursor`, and stop when `nextCursor` is null — that is the end of
+   the history. A bar still
    waiting inside its window has no record yet (`resolvesNow` shows its `closeDecision.state` as
    `DEFERRED`), and records are kept 90 days.
 4. **`get_radar_activity`** only for what steps 2 and 3 cannot do: more rows than its ten, a

package/skills/battlegrid-radar-deployment/references/radar-stage-examples.md

@@ -56,7 +56,10 @@ regime condition is the regime-reading floor, and must be a level one of the sel
 
 ## On duty only in named hours and days
 
-Hours are UTC, 0–23. Days run 0 (Sunday) to 6. Each is listed once.
+Hours are UTC, 0–23. Days run 0 (Sunday) to 6. Each is listed once. For a named exchange session —
+including the London and Tokyo opens and the London–New York overlap — stage the `hours` `get_radar_market_hours`
+serves for it rather than a set of your own; the example below is a shape, not a session. No session
+close is served: a close is staged only as hours the player names.
 
 ```json stage_radar_deployment_draft
 {