CoinRithm/coinrithm-agent-trading

Official

by CoinRithm

Overview Schema Source Related Servers Discussions Admin

TypeScript

Hybrid

Open prediction-market position

open_pm_position

Idempotent

Open a mock prediction-market position on a binary outcome with virtual mUSD; quote first, confirm, and pass a unique idempotency key.

Instructions

Open a mock prediction-market position (binary outcomes only). Requires the trade:pm scope. Enabled now (server-flag gated — returns 403 'not enabled' only if CoinRithm later disables it). idempotencyKey is REQUIRED. stakeMusd >= 10. Pass side: 'no' to back the NO side (omitted = yes); a NO entry fills at 100 minus the outcome probability and pays out if the outcome resolves false. Quote first and CONFIRM with the user. Paper trading only (virtual mUSD). Fills follow paper_execution_v1 with a disclosed execution cost; see executionModel in quote/trade results.

Input Schema

TableJSON Schema

Name	Required	Description
`side`	No	Which side of the binary outcome to back. NO pays out if it resolves false; fills at 100 minus the outcome probability. Omitted = yes.
`slug`	Yes	Prediction-market event slug.
`source`	Yes	Prediction-market source slug, e.g. kalshi or polymarket.
`thesis`	No	Optional one-line thesis for this decision (max 280 characters).
`stakeMusd`	Yes	mUSD stake (>= 10).
`agentTrace`	No	Optional private trace metadata stored in the caller's ledger.
`provenance`	No	Optional self-reported provenance (WHAT RAN). No trust: the server stamps policy versions + providerVerified itself. Any block (even {}) makes the artifact schemaVersion 2.
`idempotencyKey`	Yes	Unique per PM-open intent; reuse replays the original result.
`forecastProbability`	No	OPTIONAL. Report your OWN estimated probability (0-100, exclusive) that the chosen side wins, decided BEFORE you look at sizing/fill. It is stored SEPARATELY from the market price you pay and feeds your PUBLIC calibration record (agentBrier), which scores your forecast SKILL — not the market's. Omit it if you are not forecasting; never echo the market probability back.
`outcomeExternalMarketId`	Yes	Case-sensitive outcome or market id returned by discovery.

Output Schema

TableJSON Schema

Name	Required	Description	Default
`ok`	Yes
`body`	No
`httpStatus`	Yes
`ledgerStatus`	No
`ledgerEventId`	No

Schema Changelog

Changes observed during successful MCP inspections.

Changed5 schema fields changedv0.1.23Oct 2, 2026

removedOutput schema / properties / body / description

Removed value: -"Parsed CoinRithm response body, or raw text when the response is not JSON."

removedOutput schema / properties / httpStatus / description

Removed value: -"HTTP status returned by CoinRithm, or 0 for network errors."

removedOutput schema / properties / ledgerEventId / description

Removed value: -"Private AgentActionEvent id returned by /api/agent/*, when present."

removedOutput schema / properties / ledgerStatus / description

Removed value: -"Ledger write status header returned by CoinRithm, when present."

removedOutput schema / properties / ok / description

Removed value: -"True when CoinRithm returned a successful 2xx response."

Changed1 schema field changedv0.1.21Sep 23, 2026

addedInput schema / properties / thesis

Added value: +{
+  "description": "Optional one-line thesis for this decision (max 280 characters).",
+  "maxLength": 280,
+  "type": "string"
+}

Changed2 schema fields changedv0.1.13Jul 15, 2026

addedInput schema / properties / forecastProbability

Added value: +{
+  "description": "OPTIONAL. Report your OWN estimated probability (0-100, exclusive) that the chosen side wins, decided BEFORE you look at sizing/fill. It is stored SEPARATELY from the market price you pay and feeds your PUBLIC calibration record (agentBrier), which scores your forecast SKILL — not the market's. Omit it if you are not forecasting; never echo the market probability back.",
+  "exclusiveMaximum": 100,
+  "exclusiveMinimum": 0,
+  "type": "number"
+}

addedInput schema / properties / provenance

Added value: +{
+  "additionalProperties": false,
+  "description": "Optional self-reported provenance (WHAT RAN). No trust: the server stamps policy versions + providerVerified itself. Any block (even {}) makes the artifact schemaVersion 2.",
+  "properties": {
+    "bundleId": {
+      "maxLength": 120,
+      "type": "string"
+    },
+    "bundleVersion": {
+      "maxLength": 40,
+      "type": "string"
+    },
+    "configHash": {
+      "description": "sha256 hex of your resolved config/spec. HASH ONLY — never raw text.",
+      "pattern": "^[0-9a-fA-F]{64}$",
+      "type": "string"
+    },
+    "evidenceRef": {
+      "additionalProperties": false,
+      "description": "Pointers to the observation evidence (never the evidence itself).",
+      "properties": {
+        "snapshotIds": {
+          "description": "Opaque snapshot ids (capped at 100).",
+          "items": {
+            "maxLength": 200,
+            "type": "string"
+          },
+          "type": "array"
+        },
+        "sourceCapturedAt": {
+          "description": "Source capture time (ISO 8601).",
+          "type": "string"
+        }
+      },
+      "type": "object"
+    },
+    "modelName": {
+      "maxLength": 80,
+      "type": "string"
+    },
+    "modelProvider": {
+      "maxLength": 80,
+      "type": "string"
+    },
+    "packageVersion": {
+      "maxLength": 40,
+      "type": "string"
+    },
+    "promptHash": {
+      "description": "sha256 hex of your exact prompt strings. HASH ONLY — never raw text.",
+      "pattern": "^[0-9a-fA-F]{64}$",
+      "type": "string"
+    },
+    "runtimeKind": {
+      "description": "The runtime surface you ran on (self-reported; no trust).",
+      "enum": [
+        "hosted_scheduler",
+        "self_host_runner",
+        "byo_api",
+        "mcp_tool"
+      ],
+      "type": "string"
+    },
+    "skillVersions": {
+      "additionalProperties": {
+        "type": "string"
+      },
+      "description": "{skillId: version}. Capped: 50 keys, key<=120 / value<=40.",
+      "type": "object"
+    }
+  },
+  "type": "object"
+}

Changed1 schema field changedv0.1.12Jul 8, 2026

addedInput schema / properties / side

Added value: +{
+  "description": "Which side of the binary outcome to back. NO pays out if it resolves false; fills at 100 minus the outcome probability. Omitted = yes.",
+  "enum": [
+    "yes",
+    "no"
+  ],
+  "type": "string"
+}

Changed3 schema fields changedv0.1.8Jun 12, 2026

addedInput schema / properties / agentTrace

Added value: +{
+  "additionalProperties": false,
+  "description": "Optional private trace metadata stored in the caller's ledger.",
+  "properties": {
+    "confidence": {
+      "description": "Optional confidence score from 0 to 1.",
+      "maximum": 1,
+      "minimum": 0,
+      "type": "number"
+    },
+    "decisionId": {
+      "description": "Agent decision id for quote/write attribution.",
+      "minLength": 1,
+      "type": "string"
+    },
+    "rationaleSummary": {
+      "description": "Optional concise rationale summary. Do not include chain-of-thought, secrets, or account identity.",
+      "maxLength": 1200,
+      "minLength": 1,
+      "type": "string"
+    },
+    "runId": {
+      "description": "Agent run id for grouping.",
+      "minLength": 1,
+      "type": "string"
+    },
+    "strategyLabel": {
+      "description": "Short strategy label, self-reported by the caller.",
+      "maxLength": 120,
+      "minLength": 1,
+      "type": "string"
+    }
+  },
+  "type": "object"
+}

addedOutput schema / properties / ledgerEventId

Added value: +{
+  "description": "Private AgentActionEvent id returned by /api/agent/*, when present.",
+  "type": [
+    "string",
+    "null"
+  ]
+}

addedOutput schema / properties / ledgerStatus

Added value: +{
+  "description": "Ledger write status header returned by CoinRithm, when present.",
+  "type": [
+    "string",
+    "null"
+  ]
+}

Changed5 schema fields changedv0.1.5Jun 10, 2026

addedInput schema / properties / idempotencyKey / description

Added value: +"Unique per PM-open intent; reuse replays the original result."

addedInput schema / properties / outcomeExternalMarketId / description

Added value: +"Case-sensitive outcome or market id returned by discovery."

addedInput schema / properties / slug / description
```
Added value: +"Prediction-market event slug."
```

addedInput schema / properties / source / description

Added value: +"Prediction-market source slug, e.g. kalshi or polymarket."

changedOutput schema / (root)

Previous value: -nullNew value: +{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "additionalProperties": false,
+  "properties": {
+    "body": {
+      "description": "Parsed CoinRithm response body, or raw text when the response is not JSON."
+    },
+    "httpStatus": {
+      "description": "HTTP status returned by CoinRithm, or 0 for network errors.",
+      "type": "integer"
+    },
+    "ok": {
+      "description": "True when CoinRithm returned a successful 2xx response.",
+      "type": "boolean"
+    }
+  },
+  "required": [
+    "httpStatus",
+    "ok"
+  ],
+  "type": "object"
+}

First observedv0.1.4Jun 10, 2026

TDQS

A4.4/5.0

Behavior5/5

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

Adds substantial context beyond annotations: paper trading only (virtual mUSD), the trade:pm scope requirement, server-flag gating with a specific 403 'not enabled' failure mode, and the paper_execution_v1 fill model with a disclosed execution cost. These are exactly the behavioral traits annotations cannot convey.

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?

Front-loaded with purpose and kept to a single dense paragraph where every sentence carries operational weight. Minor redundancy with the schema on required idempotencyKey and the stake minimum, but the quote/confirm instruction earns its place.

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?

With an output schema present (executionModel in quote/trade results) and 100% schema coverage, return values need not be explained. The description still covers auth, gating, fill behavior, and safety posture, leaving no material gap for correct invocation.

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

Parameters3/5

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

Schema description coverage is 100%, so the baseline is 3. The description restates the side semantics (NO fills at 100 minus probability) and stake minimum, but these already live in the schema, so it adds little beyond what structured fields provide.

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?

States a specific verb (open) and resource (prediction-market position) with the scope qualifier 'binary outcomes only', which cleanly distinguishes it from open_futures_position and place_spot_order among siblings. An agent can identify the operation without opening the schema.

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

Usage Guidelines4/5

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

Gives a clear operational precondition ('Quote first and CONFIRM with the user') and a scope requirement, which routes the agent toward pm_quote before writing. It lacks explicit when-not conditions or named alternatives, but the sequencing guidance is actionable.

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