Skillv0.0.3

ClawScan security

Polymarket Macro Calendar Catalyst Trader · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

SuspiciousApr 28, 2026, 2:32 PM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill's code and SKILL.md largely match its trading purpose, but there are internal inconsistencies (undeclared environment variables and mismatched registry metadata) that you should understand before installing or running it with real credentials.
Guidance: This skill appears to implement the stated Polymarket catalyst-trading strategy and defaults to paper trading, which is good. However: (1) the registry metadata provided at the top is inconsistent with the files—clawhub.json and SKILL.md require SIMMER_API_KEY while the top-level 'Requirements' claimed none; (2) trader.py reads additional environment variables (thresholds and minimum trade) that are not declared in the registry metadata/tunables, so verify what environment variables the runtime will expose; (3) the only external dependency is simmer-sdk (PyPI) — review that package before installing and ensure it is the official client; (4) run the skill in sim/paper mode first and inspect network activity/requests to confirm behavior; (5) if you decide to run live, create a dedicated SIMMER_API_KEY with the least privilege necessary and rotate it afterwards. If you want higher assurance, ask the author to correct the metadata (declare all env vars and tunables) and provide a short audit of the simmer-sdk dependency and its network endpoints.

Review Dimensions

Purpose & Capability: okThe name/description (calendar-catalyst trader for Polymarket) aligns with the included code (trader.py) and SKILL.md: the skill scans markets, matches resolution dates to a catalyst calendar, computes conviction, and places trades via a SimmerClient. Required runtime dependency (simmer-sdk) and use of a SIMMER_API_KEY are appropriate for a trading skill.
Instruction Scope: noteSKILL.md and trader.py confine activity to market discovery, catalyst matching, conviction calculation, and trade placement. The skill defaults to paper trading and requires an explicit --live flag for real trades. No instructions ask the agent to read arbitrary system files or send data to unexpected external endpoints; network activity is via the Simmer SDK (expected).
Install Mechanism: okNo arbitrary download URLs or archive extraction. clawhub.json lists a pip dependency (simmer-sdk). Installing PyPI packages is a normal, moderate-risk mechanism; there are no included remote installers or opaque binaries.
Credentials: concernclawhub.json and SKILL.md explicitly require SIMMER_API_KEY (appropriate). However, trader.py reads multiple environment variables that are not declared in the top-level registry metadata: e.g., SIMMER_YES_THRESHOLD, SIMMER_NO_THRESHOLD, SIMMER_MIN_TRADE, and others are read at runtime even though only a subset of tunables are listed in clawhub.json. Additionally, the skill listing at the top of the report said "Required env vars: none", which contradicts both SKILL.md and clawhub.json. This mismatch between declared and actual env usage increases the risk of surprising behavior or unnoticed secret access.
Persistence & Privilege: okThe skill is not always:true, autostart is false, and cron is null. Automaton.managed true with entrypoint trader.py is normal for an invocable trading skill. The skill does not request system-wide config changes or elevated platform privileges in the provided files.