Vincent - Trading Engine
WarnAudited by ClawScan on May 10, 2026.
Overview
The skill is coherently an automated trading integration, but it can use trading credentials to run persistent strategies that may place real trades automatically.
Review this skill carefully before installing or enabling it on funded accounts. Its behavior is consistent with an automated trading engine, not clearly malicious, but it can keep running, consume credits, and place real trades through connected credentials. Start with canTrade disabled or very small limits, confirm approval thresholds and allowlists, and monitor active strategies and audit logs.
Findings (5)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
An active strategy may place trades automatically based on configured signals and model decisions.
The skill explicitly allows automated LLM-mediated trade decisions, which can mutate financial positions and create real monetary loss if limits, approvals, or strategy settings are wrong.
When drivers detect new information, signals are scored and batched. When the escalation threshold is met, an LLM ... decides whether to trade
Only enable trading after setting conservative spending limits, approval thresholds, allowlists, and position-sizing rules; consider disabling trade tools until the strategy is tested.
Trading automation may continue operating until paused or archived, even when the user is not actively interacting with the agent.
The skill supports persistent active strategies that keep monitoring sources and can trigger actions after the initial setup.
ACTIVE: Drivers are running. New signals trigger the pipeline.
Track which strategies are ACTIVE, pause unused strategies, and verify how to stop or archive them before enabling automation.
The skill may act through credentials that can place or manage trades on connected venues.
The trading workflow relies on venue/Vincent account credentials, which is expected but grants authority over trading accounts.
Uses the same API key as the Polymarket or HyperLiquid skill (depending on venue)
Use least-privilege credentials where available and confirm account-level trading permissions before connecting them.
Manipulated or low-quality external posts/news/signals could influence strategy decisions and lead to unwanted trades.
Untrusted external information sources can feed the model context that influences automated trading decisions.
Raw data from driver sources (web search, Twitter, newswire, price feeds, RSS, Reddit, on-chain, filings, options flow) ... The LLM evaluates batched signals against the current thesis ... issue trade decisions
Use trusted drivers, conservative thresholds, and small limits; monitor audit trails and avoid allowing unverified social/web sources to directly trigger high-value trades.
The behavior depends on the external @vincentai/cli package available at runtime.
The skill runs an external CLI package through npx; this is disclosed and purpose-aligned, but the executable package contents are not included in the reviewed artifact.
allowed-tools: Read, Write, Bash(npx:@vincentai/cli*)
Install from the expected publisher/source, prefer pinned versions when possible, and review the CLI package before using it with funded trading accounts.