# app.oddsrail/polymarket-kalshi-arbitrage (pypi · oddsrail)

Cross-venue prediction markets for AI agents: Polymarket + Kalshi, costed and attributed

- Trust score: 76/100 (medium)
- Change this week: +18
- Registry status: deprecated
- Liveness: live
- Owner verified: no
- Last scored: 2026-09-21

> **Deprecated**: this server is marked deprecated in the MCP registry. The registry records this reason: "renamed to app.oddsrail/polymarket-kalshi-trading."

## Components

- pypi · `oddsrail`: 76/100 (this document), [markdown](https://verifymcp.io/servers/app-oddsrail-polymarket-kalshi-arbitrage/oddsrail.md), [page](https://verifymcp.io/servers/app-oddsrail-polymarket-kalshi-arbitrage/oddsrail)

## Channel facts

- Registry: `pypi`
- Package: `oddsrail`
- Version: `0.10.0`
- Transport: `stdio`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-09-21.

- **Supply Chain Security**: 100/100
  - No malware found by supply-chain analysis.
  - No known CVEs affecting this package version or its production dependencies.
  - Runs hatchling.build at install time, a recognised native-build step with no shell scripting around it.
  - 1 of 58 dependencies flagged as unhealthy.
- **Provenance & Transparency**: 32/100
  - Source repository is publicly reachable at the declared URL.
  - Provenance check failed: no build-provenance attestation is published.
  - License check failed: the license (MIT License) isn't a recognized OSI-approved license.
  - Actively maintained (last published 9 days ago).
  - Disclosure check failed: no security disclosure policy was found in the source repository.
- **Schema Quality & AI Usability**: 81/100
  - 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).
  - AI-judged instruction clarity (good).
  - Tool/resource definitions use about 2656 tokens (~68/item across 39 items; 39 tools + 0 resources), lean.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 70/100
  - Stability observed for 21 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 67/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 0% of tool parameters carry a description.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - All 2 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation.
  - An AI judge read all 40 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 100/100
  - Implements a current MCP spec version (2026-07-28).

## Install

### How do I install the app.oddsrail/polymarket-kalshi-arbitrage MCP server?

app.oddsrail/polymarket-kalshi-arbitrage runs locally as a PyPI package, launched with uvx oddsrail. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

### Claude

```bash
claude mcp add app-oddsrail-polymarket-kalshi-arbitrage -- uvx oddsrail
```

### Cursor

```json
{
  "mcpServers": {
    "app-oddsrail-polymarket-kalshi-arbitrage": {
      "command": "uvx",
      "args": [
        "oddsrail"
      ]
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "app-oddsrail-polymarket-kalshi-arbitrage": {
      "command": "uvx",
      "args": [
        "oddsrail"
      ]
    }
  }
}
```

### Codex

```bash
codex mcp add app-oddsrail-polymarket-kalshi-arbitrage -- uvx oddsrail
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "app-oddsrail-polymarket-kalshi-arbitrage": {
      "type": "local",
      "command": [
        "uvx",
        "oddsrail"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add app-oddsrail-polymarket-kalshi-arbitrage --command uvx --arg oddsrail
```

### Hermes

```yaml
mcp_servers:
  app-oddsrail-polymarket-kalshi-arbitrage:
    command: "uvx"
    args: ["oddsrail"]
```

### Netclaw

```json
{
  "McpServers": {
    "app-oddsrail-polymarket-kalshi-arbitrage": {
      "Transport": "stdio",
      "Command": "uvx",
      "Arguments": [
        "oddsrail"
      ]
    }
  }
}
```

### Vellum

```bash
assistant mcp add app-oddsrail-polymarket-kalshi-arbitrage -t stdio -c uvx -a oddsrail
```

### Other

```json
{
  "mcpServers": {
    "app-oddsrail-polymarket-kalshi-arbitrage": {
      "command": "uvx",
      "args": [
        "oddsrail"
      ]
    }
  }
}
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-09-20 (score 76, +1)

No change was recorded against any check on this day. Stability & Change Management went from 63 to 67. That category is still filling its 30-day observation window: 19 days of observed history at the previous scan, 20 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-09-18 (score 75, +16)

- [security improvement] Malware scan: unverified → pass

### 2026-09-17 (score 59, −15)

- [security regression] Malware scan: pass → unverified

### 2026-09-16 (score 74, +3)

- [security improvement] Known CVEs: partial → pass
- [functional] Dependency health: partial → 1.00

### 2026-09-15 (score 71, +13)

- [security regression] Known CVEs: pass → partial
- [security improvement] Malware scan: unverified → pass
- [functional] Dependency health: 1.00 → partial

### 2026-09-14 (score 58, −14)

- [security regression] Malware scan: pass → unverified

### 2026-09-11 (score 72, +3)

- [security improvement] Known CVEs: partial → pass
- [functional] Dependency health: partial → 1.00

### 2026-09-10 (score 69, −2)

- [security regression] Known CVEs: pass → partial
- [functional] Dependency health: 1.00 → partial

## MCP tools (39)

### `search_markets` (~55 tokens)

Search Polymarket markets by text (Gamma public-search under the hood); empty query lists open markets. Returns token ids, prices, metrics, resolution info.

Input parameters:

- `limit` (integer)
- `query` (string)

### `get_market` (~28 tokens)

Get one market's details by slug (or id).

Input parameters:

- `id_or_slug` (string, required)

### `get_orderbook` (~35 tokens)

Get the live orderbook (bids/asks) for a CLOB token id.

Input parameters:

- `token_id` (string, required)

### `price_history` (~49 tokens)

Recent price history for a CLOB token id: hours back, at fidelity_minutes resolution.

Input parameters:

- `fidelity_minutes` (integer)
- `hours` (number)
- `token_id` (string, required)

### `get_positions` (~29 tokens)

Current positions for a wallet address.

Input parameters:

- `address` (string, required)
- `limit` (integer)

### `overshoot_signal` (~79 tokens)

PREMIUM SIGNAL — overshoot/fade detector. Analyzes a token's recent price series for fresh panic jumps and reports whether a fade setup is active plus this market's historical reversion tendency.

Input parameters:

- `hours` (number)
- `lookback_s` (number)
- `threshold` (number)
- `token_id` (string, required)

### `dispute_risk` (~48 tokens)

PREMIUM SIGNAL — dispute-risk triage. Scores 0-100 how likely a market's resolution gets contested (UMA dispute risk) with transparent reasons.

Input parameters:

- `id_or_slug` (string, required)

### `place_order` (~145 tokens)

Place a limit order. DRY-RUN by default: returns the order it would post. Real trading needs ODDSRAIL_DRY_RUN=0 and POLYMARKET_PRIVATE_KEY. The operator's builder code is signed into the order. price is the implied probability in (0,1); size is in SHARES (notional = price * size), and the exchange enforces a $1 minimum notional on marketable orders. post_only=True rejects rather than crosses the book.

Input parameters:

- `post_only` (boolean)
- `price` (number, required)
- `side` (string, required)
- `size` (number, required)
- `token_id` (string, required)

### `cancel_order` (~28 tokens)

Cancel an open order by id (respects dry-run).

Input parameters:

- `order_id` (string, required)

### `open_orders` (~16 tokens)

List the operator wallet's open orders.

### `builder_stats` (~48 tokens)

Builder attribution stats: the public builder leaderboard, and (if ODDSRAIL_BUILDER_CODE is set) matched trades attributed to this operator's code.

Input parameters:

- `time_period` (string)

### `kalshi_search_markets` (~75 tokens)

Search Kalshi markets. Kalshi has no text-search endpoint, so this pages open markets and filters on title/ticker; auto-generated MVE combo shards are excluded. Prices are dollar strings, not cents.

Input parameters:

- `limit` (integer)
- `min_volume` (number)
- `query` (string)

### `kalshi_get_market` (~25 tokens)

Get one Kalshi market by ticker.

Input parameters:

- `ticker` (string, required)

### `kalshi_get_orderbook` (~67 tokens)

Kalshi orderbook for a ticker, normalised to a YES-book bid/ask view (Kalshi publishes bid ladders only; asks are derived as 1 - NO bid). Raw ladders included.

Input parameters:

- `depth` (integer)
- `ticker` (string, required)

### `kalshi_get_trades` (~34 tokens)

Recent public trades for a Kalshi ticker.

Input parameters:

- `limit` (integer)
- `ticker` (string, required)

### `kalshi_balance` (~21 tokens)

Kalshi account balance (needs the operator's API key).

### `kalshi_positions` (~27 tokens)

Kalshi positions (needs the operator's API key).

Input parameters:

- `limit` (integer)

### `kalshi_open_orders` (~29 tokens)

Kalshi resting orders (needs the operator's API key).

Input parameters:

- `limit` (integer)

### `kalshi_place_order` (~107 tokens)

Place a Kalshi limit order. State it naturally: outcome yes|no, action buy|sell, price = probability of THAT outcome in (0,1). Translated to Kalshi's YES-book bid/ask internally. DRY-RUN by default.

Input parameters:

- `action` (string, required)
- `count` (number, required)
- `outcome` (string, required)
- `price` (number, required)
- `ticker` (string, required)
- `time_in_force` (string)

### `kalshi_cancel_order` (~31 tokens)

Cancel a Kalshi order by id (respects dry-run).

Input parameters:

- `order_id` (string, required)

### `find_markets` (~110 tokens)

Search BOTH Polymarket and Kalshi at once and return one normalised shape per market: venue, market_id (the id that venue's place_order takes), title, yes/no price as probabilities in (0,1), best bid/ask, spread, 24h volume, close time. Use this instead of the per-venue search tools when you do not already know the venue.

Input parameters:

- `limit` (integer)
- `query` (string, required)
- `venues` (string)

### `compare_venues` (~112 tokens)

Find markets that may be the SAME event on both Polymarket and Kalshi. NOT an arbitrage scanner: matches are candidates from title similarity plus a close-date check, and a price difference between two candidates is not profit. Identical wording does not mean identical resolution criteria — read both, and run quote_cost on each leg, before acting.

Input parameters:

- `limit` (integer)
- `max_close_days_apart` (integer)
- `min_similarity` (number)
- `query` (string, required)

### `quote_cost` (~108 tokens)

What a given size would ACTUALLY cost, by walking the order book rather than reading the top level. Returns average fill price, slippage vs best, notional, and the levels consumed — plus the venue's fee schedule where it publishes one. Call this before sizing any trade, and on both legs before acting on a cross-venue gap.

Input parameters:

- `market_id` (string, required)
- `side` (string, required)
- `size` (number, required)
- `venue` (string, required)

### `order_status` (~44 tokens)

Status of one Polymarket order: resting, partially filled, filled, or gone. The lifecycle answer an agent needs after place_order.

Input parameters:

- `order_id` (string, required)

### `my_fills` (~35 tokens)

Recent executions (fills) for the operator wallet — confirms what actually traded, with tx hashes.

Input parameters:

- `limit` (integer)

### `my_positions` (~33 tokens)

The operator's current Polymarket positions (uses the configured wallet; no address needed).

Input parameters:

- `limit` (integer)

### `cancel_all_orders` (~41 tokens)

KILL SWITCH — cancel every resting Polymarket order on the operator account at once. Use when exposure must go to zero fast. Respects dry-run.

### `split_position` (~114 tokens)

Split USDC collateral into a full YES+NO share set for one market, GASLESS via Polymarket's relayer. amount_usdc is collateral in USDC (e.g. 25). Needs POLYMARKET_RELAYER_API_KEY and POLYMARKET_RELAYER_API_KEY_ADDRESS (polymarket.com -> Settings -> Relayer API keys). DRY-RUN by default. Never falls back to a gas-paying broadcast.

Input parameters:

- `amount_usdc` (number, required)
- `condition_id` (string, required)

### `merge_positions` (~72 tokens)

Merge matching YES+NO shares back into USDC, GASLESS via the relayer. amount is in shares, or 'max' for the largest balanced amount held. Needs the operator's relayer key. DRY-RUN by default.

Input parameters:

- `amount` (string)
- `condition_id` (string, required)

### `redeem_positions` (~71 tokens)

Redeem the winning shares of a RESOLVED market for USDC, GASLESS via the relayer. Pass exactly one of condition_id or market_id. Needs the operator's relayer key. DRY-RUN by default.

Input parameters:

- `condition_id` (string)
- `market_id` (string)

### `paper_positions` (~78 tokens)

Paper-trading portfolio for dry-run: cash, positions at current marks, realized and unrealized P&L, resting paper orders (filled here if the market has crossed them). Dry-run Polymarket orders are papered against the live book automatically. Simulated: no queue, no impact, no fees, so results are an upper bound.

### `paper_reset` (~45 tokens)

Reset the paper-trading ledger to its starting bankroll (ODDSRAIL_PAPER_BANKROLL, default 1000 USDC). Deletes simulated fills and positions; touches nothing real.

### `watch_book` (~96 tokens)

Stream one Polymarket token's realtime market events (book snapshot, price changes, trades) for up to `seconds` (1-60) or `max_events`, then return them. Use after get_orderbook when you need to see the book MOVE before acting; a quiet market may deliver only the initial snapshot.

Input parameters:

- `max_events` (integer)
- `seconds` (number)
- `token_id` (string, required)

### `redeemable_positions` (~61 tokens)

What the operator can turn back into USDC now: redeemable (resolved, winning) positions and mergeable (hold both YES and NO) positions, with the ids the gasless tools take. Uses the configured wallet.

Input parameters:

- `limit` (integer)

### `resolution_criteria` (~74 tokens)

READ BEFORE TRUSTING A PRICE: the full resolution contract for a market — what exactly resolves YES, who resolves it, from which sources. venue is 'polymarket' (pass slug or id) or 'kalshi' (pass ticker).

Input parameters:

- `market_id` (string, required)
- `venue` (string, required)

### `closing_soon` (~47 tokens)

Markets closing within N hours on either venue, by volume — where trading activity concentrates.

Input parameters:

- `hours` (number)
- `limit` (integer)
- `venues` (string)

### `settlement_audit` (~100 tokens)

Settlement-divergence audit for a cross-venue pair, on LIVE data (no pre-curated pair list). Compares close times, resolution sources, UMA dispute status and market structure, and returns ok / caution / block with reasons. Run this before treating any cross-venue price difference as an edge.

Input parameters:

- `kalshi_ticker` (string, required)
- `notional_usd` (number)
- `polymarket_id` (string, required)

### `position_size` (~93 tokens)

Fractional-Kelly position size for a binary contract, given your bankroll, the market price, and YOUR fair value estimate. Caps at a fraction of full Kelly and refuses negative-edge bets. Returns its assumptions — subtract quote_cost before trusting the number.

Input parameters:

- `bankroll_usd` (number, required)
- `fair_value` (number, required)
- `max_fraction_of_kelly` (number)
- `price` (number, required)

### `server_info` (~62 tokens)

Server status: dry-run state, attribution config, enabled capabilities, venue reachability from this machine, and Polymarket's geoblock verdict for this machine's IP (advisory — not a compliance check). Call this before the first order of a session.

## Diagnostics

Captured diagnostic sections: Provenance, Install scripts, Dependencies. The full working is on the page: https://verifymcp.io/servers/app-oddsrail-polymarket-kalshi-arbitrage/oddsrail#diagnostics

## Score history

- 2026-09-21: 76
- 2026-09-20: 76
- 2026-09-19: 75
- 2026-09-18: 75
- 2026-09-17: 59
- 2026-09-16: 74
- 2026-09-15: 71
- 2026-09-14: 58
- 2026-09-13: 72
- 2026-09-12: 72
- 2026-09-11: 72
- 2026-09-10: 69
- 2026-09-09: 71
- 2026-09-08: 70
- 2026-09-07: 70
- 2026-09-06: 69
- 2026-09-05: 69
- 2026-09-04: 68
- 2026-09-03: 68
- 2026-09-02: 52
- 2026-09-01: 67
- 2026-08-31: 66

## Common questions

### What is the app.oddsrail/polymarket-kalshi-arbitrage MCP server?

app.oddsrail/polymarket-kalshi-arbitrage is an MCP server listed in the public MCP registry as app.oddsrail/polymarket-kalshi-arbitrage. Cross-venue prediction markets for AI agents: Polymarket + Kalshi, costed and attributed. This page covers its PyPI package (oddsrail).

### Is the app.oddsrail/polymarket-kalshi-arbitrage MCP server safe to use?

app.oddsrail/polymarket-kalshi-arbitrage scores 76 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 21 September 2026. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

### What tools does the app.oddsrail/polymarket-kalshi-arbitrage MCP server expose?

app.oddsrail/polymarket-kalshi-arbitrage exposes 39 tools: search_markets, get_market, get_orderbook, price_history, get_positions, and 34 more. Their descriptions and schemas cost roughly 2,372 tokens of context every time the server is loaded.

### Is the app.oddsrail/polymarket-kalshi-arbitrage MCP server still maintained?

app.oddsrail/polymarket-kalshi-arbitrage is marked deprecated in the MCP registry. We last reached this channel on 21 September 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

## Links

- PyPI project: https://pypi.org/project/oddsrail/
- Socket report: https://socket.dev/pypi/package/oddsrail
- Repository: https://github.com/hmesutozsoy/oddsrail
- Changelog RSS feed: https://verifymcp.io/servers/app-oddsrail-polymarket-kalshi-arbitrage/oddsrail.xml
- Changelog JSON feed: https://verifymcp.io/servers/app-oddsrail-polymarket-kalshi-arbitrage/oddsrail.json
- HTML version of this page: https://verifymcp.io/servers/app-oddsrail-polymarket-kalshi-arbitrage/oddsrail