Polymarket Macro Sentiment Divergence Trader

Detects macro sentiment divergence across Polymarket prediction markets. When positive-sentiment categories (sports winners, tech milestones, entertainment,...

Security Scan

The skill's code, instructions, and declared requirements are consistent with a Polymarket/Simmer trading strategy: it only needs a Simmer API key, defaults to paper trading, and its behavior matches the SKILL.md and clawhub.json.

Purpose & Capability

Name and description describe a market-divergence trading strategy; the code implements market discovery, classification, divergence scoring, and trading via the simmer-sdk. The single required credential (SIMMER_API_KEY) and the declared pip dependency (simmer-sdk) are appropriate for the stated purpose.

Instruction Scope

SKILL.md instructs the agent to scan markets, classify sentiment, compute a divergence metric, and (optionally) place trades; the code follows those steps. The skill explicitly defaults to paper trading and only executes real trades with an explicit --live flag. The instructions do not request unrelated files or system credentials.

Install Mechanism

There is no complex custom install script; however clawhub.json declares a pip dependency on 'simmer-sdk' which will be installed from PyPI. Installing third-party Python packages is normal for this use case but carries the usual supply-chain risk — verify the package and maintainer if you want extra assurance.

Credentials

Only SIMMER_API_KEY is required by the skill (plus optional tunables via environment). That credential is necessary and proportionate for placing trades through the Simmer client. No unrelated secrets or system paths are requested.

Persistence & Privilege

autostart is false and cron is null; the skill will not run automatically by default. always:true is not set. The skill declares an automaton entrypoint but does not force permanent presence or modify other skills' configuration. Agent autonomous invocation remains possible (platform default) but is not an additional red flag here.

Assessment

This skill appears to do what it says: analyze Polymarket-style markets and trade via the Simmer API. Before installing, make sure you: 1) Understand that SIMMER_API_KEY is a high-value credential — only provide keys with minimum required permissions and rotate them if you stop using the skill. 2) Run the skill in paper/sim mode first (default) to validate behavior and tune parameters. 3) Review the simmer-sdk PyPI package and its GitHub repo to ensure you trust the package source. 4) Be aware the sentiment classifier is keyword-based and may misclassify markets; monitor live runs closely and limit max position/tunables. 5) Avoid running with --live until you’ve tested it thoroughly and are comfortable with financial risk.

Macro Sentiment Divergence Trader

This is a template. The default signal classifies markets into positive/negative sentiment buckets, computes a divergence index, and trades against the stale side when macro sentiment is logically inconsistent. The skill handles all the plumbing (market discovery, trade execution, safeguards). Your agent provides the alpha.

Strategy Overview

Prediction markets collectively reflect macro sentiment. "Positive" markets (sports favorites winning, tech milestones being hit, crypto going UP, entertainment events succeeding) and "negative" markets (geopolitical escalation, catastrophe events, disease outbreaks, crypto going DOWN) should not both be priced high simultaneously. If the average probability of positive-sentiment markets is 0.65 and the average probability of negative-sentiment markets is 0.60, both sides are saying "likely" -- but the world cannot be both great and terrible at the same time. One side is stale.

The structural edge: retail traders price individual markets in isolation. They don't cross-reference whether their collective bets are internally consistent. A sports fan pricing "Lakers win championship" at 70% is not checking whether "major recession in 2026" is also at 65%. But those two bets are in tension -- championship attendance, sponsorship revenue, and betting volumes all correlate with economic health.

The divergence formula: positive_index + negative_index - 1.0. When both indices exceed 0.5, the sum exceeds 1.0 and the divergence is positive. A divergence of 0.15+ means the collective market is pricing in logically contradictory macro states.

Signal Logic

Default Signal: Sentiment Divergence Detection

1. Scan ALL active markets via get_markets(limit=200) plus targeted keyword search
2. Classify each market as positive-sentiment, negative-sentiment, or unclassified
3. Compute positive_index = average probability of positive-sentiment markets
4. Compute negative_index = average probability of negative-sentiment markets
5. divergence = positive_index + negative_index - 1.0
6. If divergence > SIMMER_DIVERGENCE_THRESHOLD (default 0.15):
   • Identify which side has LESS recent movement (the "stale" side)
   • Trade AGAINST the stale side -- those markets haven't repriced to new information
7. Conviction-based sizing with divergence magnitude as a boost factor

Sentiment Classification

Positive sentiment (things going well):

• Sports favorites winning (champion, victory, gold medal, Super Bowl, finals)
• Tech milestones (launch, milestone, breakthrough, IPO, market cap, trillion)
• Entertainment success (box office, Oscar, Grammy, streaming record)
• Crypto UP (bitcoin above, btc above, new all-time high)

Negative sentiment (things going badly):

• Geopolitical escalation (war, invasion, conflict, sanctions, military, nuclear)
• Catastrophe events (hurricane, earthquake, tsunami, wildfire, flood, famine)
• Disease outbreaks (pandemic, epidemic, outbreak, bird flu, covid)
• Economic doom (recession, crash, collapse, default, bankruptcy)
• Crypto DOWN (bitcoin below, btc below)

Stale Side Detection

The side with less recent price movement (closer to 50% on average) is considered stale -- it hasn't repriced to whatever new information caused the other side to move. Distance from 50% serves as a proxy for recency of directional movement.

Why This Works

• Retail prices individual markets without cross-referencing macro consistency
• Positive and negative macro states are inversely correlated in reality
• When both sides are priced high, at least one side is wrong
• The stale side (less recent movement) is the one that hasn't absorbed new info
• Mean reversion of the stale side toward consistency is the trade

Safety & Execution Mode

The skill defaults to paper trading (venue="sim"). Real trades only with --live flag.

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

autostart: false and cron: null -- nothing runs automatically until you configure it in Simmer UI.

Required Credentials

Variable	Required	Notes
SIMMER_API_KEY	Yes	Trading authority. Treat as high-value credential.

Tunables (Risk Parameters)

All declared as tunables in clawhub.json and adjustable from the Simmer UI.

Variable	Default	Purpose
SIMMER_MAX_POSITION	40	Max USDC per trade (reached at 100% conviction)
SIMMER_MIN_VOLUME	5000	Min market volume filter (USD)
SIMMER_MAX_SPREAD	0.10	Max bid-ask spread (10%)
SIMMER_MIN_DAYS	3	Min days until resolution
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_TRADE	5	Floor for any trade (min USDC regardless of conviction)
SIMMER_DIVERGENCE_THRESHOLD	0.15	Min divergence score to trigger trades
SIMMER_MIN_BUCKET_SIZE	3	Min markets required per sentiment bucket

Dependency

simmer-sdk by Simmer Markets (SpartanLabsXyz)

• PyPI: https://pypi.org/project/simmer-sdk/
• GitHub: https://github.com/SpartanLabsXyz/simmer-sdk