Polymarket Copy Consensus Trader

Only trades Polymarket markets where 3+ top whale wallets independently agree on direction. Filters out noisy single-wallet signals and focuses on high-convi...

Security Scan

The skill's code, declared dependencies, and required credential are consistent with its stated purpose (consensus-based Polymarket trading); nothing in the package suggests hidden exfiltration or unrelated privileges, but it will make outbound requests and can place live trades when given the SIMMER_API_KEY and run with --live.

Purpose & Capability

Name/description match the implementation: the code fetches a leaderboard from predicting.top, pulls wallet activity from data-api.polymarket.com, builds a consensus map, and uses the Simmer SDK to place simulated or live trades. The only required credential (SIMMER_API_KEY) and the declared pip dependency (simmer-sdk) are appropriate for a trading skill.

Instruction Scope

SKILL.md and trader.py stay within the stated scope: they describe and implement fetching public leaderboards and wallet activity, building consensus, and trading via the Simmer SDK. Runtime instructions and code do not read unrelated local files or request unrelated environment variables. They do perform external network calls to the public endpoints and to the Simmer backend (via simmer-sdk), which is expected for this skill.

Install Mechanism

No explicit install script is included, but clawhub.json declares a pip requirement on 'simmer-sdk'. Installing a third-party pip package is a normal moderate-risk step for this kind of skill; you should verify the simmer-sdk package's provenance before installation (PyPI/org source) if you are concerned.

Credentials

Only SIMMER_API_KEY is required (declared in clawhub.json and SKILL.md). That single credential is proportional to a trading skill that needs an API key to place orders. Other tunables are configurable via SIMMER_ env vars and are reasonable for risk parameters.

Persistence & Privilege

The skill is not always-on (always:false), autostart is false, and the automaton cron is null by default. It does call apply_skill_config() when available but does not modify other skills or system-wide settings. Autonomous invocation is enabled by default on the platform, which is normal; there is no evidence this skill requests elevated permanence beyond that.

Assessment

This package appears to do what it claims: it will fetch public data (predicting.top and Polymarket), aggregate wallet signals, and use the simmer-sdk client to paper-trade by default or place real trades if you run it with --live and provide SIMMER_API_KEY. Before installing or running live: 1) only provide SIMMER_API_KEY if you trust the Simmer service and understand the key's permissions (it can place real USDC trades); 2) verify the simmer-sdk package source/version on PyPI or your org's registry; 3) keep autostart disabled and test in paper mode to validate behavior; 4) review the code if you need assurance about how many external endpoints are contacted and what data is sent to the Simmer backend. If you plan to run autonomously, consider monitoring logs and limiting the API key's privileges or funds available to reduce risk.

Copy Consensus Trader

This is a template. The default signal requires multi-wallet consensus before trading -- remix it with SmartScore weighting, time decay on activity, or sector-specific consensus thresholds. The skill handles all the plumbing (leaderboard fetching, activity aggregation, consensus building, conviction-boosted sizing, trade execution, safeguards). Your agent provides the alpha.

Strategy Overview

A single whale can be wrong, manipulating, or hedging a larger position elsewhere. But when 3+ independent smart-money wallets all bet the same direction on the same market, the signal is fundamentally different. This skill fetches top wallets from the public predicting.top leaderboard, pulls each wallet's recent activity from the Polymarket data API, and builds a consensus map aggregating net positions across all wallets per market.

Only markets where the majority side has at least 3 wallets AND opposition is minimal (fewer than half the consensus count) qualify as tradeable signals. This strict filter eliminates noisy single-wallet signals and focuses capital on the highest-conviction setups where smart money independently agrees.

Signal Logic

Step 1: Fetch Leaderboard

Top wallets are fetched from predicting.top/api/leaderboard. The skill tracks up to 20 wallets (configurable via SIMMER_LEADERBOARD_LIMIT).

Step 2: Fetch Wallet Activity

For each wallet, recent trades (up to 80) are fetched from the Polymarket data API (data-api.polymarket.com). Net positions are extracted per market: if a wallet has more YES volume than NO volume on a market, that counts as a YES position.

Step 3: Build Consensus Map

All wallet positions are aggregated into a consensus map:

market_title -> {yes_wallets: set, no_wallets: set, yes_volume, no_volume}

Step 4: Find Consensus Signals

Markets are filtered to those where:

• The majority side has >= MIN_CONSENSUS wallets (default: 3)
• Opposition is clean: either zero opposing wallets OR opposing < consensus / 2

Signals are sorted by wallet count descending, then total volume descending.

Step 5: Trade with Consensus Boost

For each consensus signal:

1. Find matching Simmer market via progressive keyword search
2. Run standard compute_signal() -- conviction-based sizing per CLAUDE.md
3. Only trade if consensus direction matches signal direction (dual confirmation)
4. Apply consensus boost: extra_whales * CONSENSUS_BOOST (capped at 50%)
5. Boosted conviction = base_conviction * (1 + boost), capped at 1.0
6. Check context_ok() for flip-flop and slippage safeguards

Conviction Sizing (base, before consensus boost)

p	conviction	size
38%	0%	$5 (floor)
30%	21%	$10.50
20%	47%	$23.50
0%	100%	$50

Consensus Boost Example

With 5 wallets agreeing (MIN_CONSENSUS=3, CONSENSUS_BOOST=0.20):

• Extra whales: 5 - 3 = 2
• Boost: 2 * 0.20 = 0.40 (40% conviction increase)
• Base conviction 0.47 becomes min(1.0, 0.47 * 1.40) = 0.658
• Size: max(5, round(0.658 * 50, 2)) = $32.90

Edge Thesis

The core edge is that independent agreement among multiple smart-money wallets is a much stronger signal than any single wallet's position:

1. Noise reduction: Single wallets can be hedging, testing, or making mistakes. Multiple wallets agreeing independently filters all of that out.
2. Information aggregation: Each whale may have access to different information sources. Consensus means multiple independent information channels point the same way.
3. Reduced manipulation risk: One whale can manipulate a market. Coordinating 3+ independent wallets to all manipulate the same market in the same direction is far harder.
4. Conviction scaling: The more whales that agree, the stronger the position sizing -- reflecting the exponentially decreasing probability that all of them are wrong.

Remix Ideas

• SmartScore weighting: Weight each wallet's vote by their SmartScore -- a 95-score whale counts more than a 70-score whale
• Time decay: Weight recent activity (last 24h) more heavily than older positions (last 7d)
• Sector filtering: Only count consensus within categories each whale historically performs best in
• Anti-correlation: Track which wallets tend to disagree -- if historically uncorrelated wallets suddenly agree, that is an even stronger signal
• Dynamic thresholds: Raise MIN_CONSENSUS during high-volatility regimes, lower it during quiet periods

Safety & Execution Mode

The skill defaults to paper trading (venue="sim"). Real trades only execute when --live is passed explicitly.

Scenario	Mode	Financial risk
python trader.py	Paper (sim)	None
Cron / automaton	Paper (sim)	None
python trader.py --live	Live (polymarket)	Real USDC

The automaton cron is set to null -- it does not run on a schedule until you configure it in the Simmer UI. autostart: false means it won't start automatically on install.

Required Credentials

Variable	Required	Notes
SIMMER_API_KEY	Yes	Trading authority -- keep this credential private.

Tunables (Risk Parameters)

All risk parameters are declared in clawhub.json as tunables and adjustable from the Simmer UI without code changes. They use SIMMER_-prefixed env vars so apply_skill_config() can load them securely.

Variable	Default	Purpose
SIMMER_MAX_POSITION	50	Max USDC per trade (reached at 100% conviction)
SIMMER_MIN_TRADE	5	Floor for any trade (min USDC regardless of conviction)
SIMMER_MIN_VOLUME	5000	Min market volume filter (USD)
SIMMER_MAX_SPREAD	0.08	Max bid-ask spread (0.08 = 8%)
SIMMER_MIN_DAYS	5	Min days until market resolves
SIMMER_MAX_POSITIONS	8	Max concurrent open positions
SIMMER_YES_THRESHOLD	0.38	Buy YES if market price <= this value
SIMMER_NO_THRESHOLD	0.62	Sell NO if market price >= this value
SIMMER_MIN_CONSENSUS	3	Min wallets that must agree for a consensus signal
SIMMER_CONSENSUS_BOOST	0.20	Extra conviction per whale beyond minimum (capped at 0.50 total)
SIMMER_LEADERBOARD_LIMIT	20	How many top wallets to track from the leaderboard

Dependency

simmer-sdk is published on PyPI by Simmer Markets.

• PyPI: https://pypi.org/project/simmer-sdk/
• GitHub: https://github.com/SpartanLabsXyz/simmer-sdk
• Publisher: [e-mail]

Review the source before providing live credentials if you require full auditability.