@battlegrid/mcp-server 31.2.15 → 31.2.17

package/README.md

@@ -24,7 +24,7 @@ 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 — v37 → v53
+## Contract history — v37 → v54
 
 Eleven majors reached authors while this section stopped at v36. That gap is the mechanism, not an
 oversight: since v31 a contract move needs no release here, so nothing forced a note to be written —
@@ -291,6 +291,45 @@ refuse is now stored. Listed because a client that special-cased the refusal can
 
 ### Reshaped output — the same call returns a different shape
 
+- **`scan_agent_coins` returns the ranking, not an explanation of every coin** (54.0.0,
+  `fix-mcp-scan-row-altitude`). The scan used to wrap the app's full per-coin qualification verdict
+  in every row — two directions with candidate-level construction and a stop-loss policy band, four
+  gates each with its own measurement, condition reach reasons, the ATR corridor. For a 78-coin
+  catalog that was **82,147 characters, ~1,053 per row**, which is past the tool-result cap of every
+  client we know of: the calling model received a file path instead of an answer, so the tool did not
+  deliver its result even when the scan succeeded.
+
+  `rows` is gone. Three ranked arrays replace it, and `rank` is **global across all three**, so
+  reading them in this order reproduces the server's own sequence:
+
+  ```
+  qualified[]   { rank, coinTicker, scorePercent, coinDataStopped }
+  rejected[]    { rank, coinTicker, scorePercent, firstFailReason, scoreShortfallPercent }
+  unscorable[]  { rank, coinTicker, coinDataStopped }
+  ```
+
+  Array membership now carries what the row `kind` discriminator and the `qualifies` flag used to,
+  and both are gone with them; `firstFailReason` is non-nullable on a rejected row, because a
+  non-qualifying verdict always names the gate that blocked it. Rows also lose `long`, `short`,
+  `gates`, `tradeableAtrRange`, `evaluatedAt`, `coinName`, `assetClass` and `category`.
+
+  Two things are new. The agent's `agentId`, `agentName`, `strategyTimeframe` and `minScorePercent`
+  move to an `agent` object carried **once** instead of on all 78 rows — `null` when the scan scored
+  nothing at all. And `scoreShortfallPercent` is server-computed: how far below the minimum the score
+  fell, non-null exactly when the aggregate score is what blocked, so you never subtract a published
+  threshold from a published reading yourself.
+
+  `scanStartedAt`, `coinsScanned` and `qualifiedCount` are unchanged, and so is every coin: same
+  rows, same order, same ranks, same verdicts, same rate buckets, same evaluator. Only the fields
+  moved — about 8.8 KB for the same 78-coin scan.
+
+  **Migration.** Read `qualified` / `rejected` / `unscorable` instead of `rows`, and take the agent's
+  thresholds from `agent` rather than from the first row. For a shortlisted coin's full per-direction
+  and per-gate detail, call `get_agent_coin_qualification` on up to 12 tickers — it carries every
+  dropped field and costs no second scan (the scan is rate-limited to 3 per agent and 10 per user a
+  minute; the probe is not).
+
+
 - **Report headers, glosses and signal indicator keys are renamed for the Donchian channel**
   (53.0.0, `rename-donchian-channel`). Every report surface — `preview_strategy_report`,
   `get_strategy_section_template`, the agent prompt previews — renders `donchianHi` / `donchianLo`

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 = "31.2.15";
+export declare const PACKAGE_VERSION = "31.2.17";
 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 = '31.2.15';
+export const PACKAGE_VERSION = '31.2.17';
 export const DEFAULT_URL = 'https://mcp.battlegrid.trade/mcp';
 const MAX_RETRIES = 3;
 const RETRY_DELAYS_MS = [2000, 4000, 8000];

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@battlegrid/mcp-server",
-  "version": "31.2.15",
-  "description": "BattleGrid MCP server — play crypto prediction games from AI agents",
+  "version": "31.2.17",
+  "description": "BattleGrid MCP server \u2014 play crypto prediction games from AI agents",
   "type": "module",
   "main": "dist/index.js",
   "bin": {

package/skills/EXPORT.json

@@ -1,6 +1,6 @@
 {
   "generator": "battlegrid-app/server/scripts/export-mcp-skills.mjs",
-  "contractVersion": "53.0.0",
+  "contractVersion": "54.0.0",
   "files": {
     "battlegrid-agent-management/SKILL.md": "b6ea75b3d838c1dbaaf3984d7e99a11c81686de2c1a1910c35699ee7e70dafa5",
     "battlegrid-arena-play/SKILL.md": "03fa153bf82be18bf5ed01e3ba12e2cd9f99a81b08d903a77b52169e7519a182",
@@ -13,6 +13,6 @@
     "battlegrid-strategy-examples/references/recipes.md": "53887daece17995543e7d3fe1bf2577fc41a6254c291acacde48d7de63853641",
     "battlegrid-strategy-examples/references/tradingview-ports.md": "1ff63367de1f9a9fbbcd7ecd5367137800373df47aaeb1d61fb1298d3a1924f4",
     "battlegrid-trade-analysis/SKILL.md": "32573a67e55d735e6ee4e36b3b48c815694102704bc453c4d1c13db39551b3da",
-    "battlegrid-trade-proposal/SKILL.md": "0a4586f660f605281c7775523c9e41422c6e52646ce52bf0688254af4619e46c"
+    "battlegrid-trade-proposal/SKILL.md": "311af5a91576b095dcfcb3e3fc8b41e8a75d7af505222f975fde9733c633cb7e"
   }
 }

package/skills/battlegrid-trade-proposal/SKILL.md

@@ -23,6 +23,9 @@ re-sorts a list, re-derives a verdict, or approves on the player's behalf.
 - Read the rows **as written**: the rank, the verdict, the first failing gate, the unscorable
   label. Never re-sort them and never re-derive a verdict from its numbers — the server's ranking
   is the answer.
+- The scan is the **ranking** answer, not the full gate detail. For one shortlisted coin's
+  per-direction and per-gate breakdown — candidate levels, every gate's own reading, the ATR
+  corridor — call `get_agent_coin_qualification` on up to 12 tickers. Never re-scan for it.
 - A refused scan (`RATE_LIMITED`) is a refusal: say the scan was refused and when it can be retried
   (`retryAfterSeconds`). Never say "nothing fits".