Xerolite
WarnAudited by ClawScan on May 10, 2026.
Overview
The skill appears coherent, but it lets an agent place brokerage orders with an API key and does not document confirmation, limit, or safety controls.
Use this only with a trusted Xerolite/IBKR setup. Prefer paper trading or tightly scoped credentials, require explicit confirmation before every order, verify order details carefully, and avoid exposing API keys or portfolio output in shared logs or transcripts.
Findings (3)
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.
A mistaken, ambiguous, or unintended agent action could submit real trades and affect the user's brokerage account.
The skill explicitly enables automated trade execution, but the artifacts do not describe confirmation prompts, order limits, paper-trading mode, or other containment for real financial orders.
orders are sent in real time with no manual steps ... call the Xerolite REST API to **place orders** ... from natural language or automation
Require explicit user confirmation before any order placement, verify symbol/action/quantity, use paper trading or small limits where possible, and consider separating read-only portfolio/search access from trading access.
Anyone or any agent process with access to this key may be able to use the configured Xerolite trading and portfolio endpoints.
The CLI authenticates with a local API key and sends it to the Xerolite API, which is expected for the stated purpose but grants privileged account access.
const apiKey = flags['api-key'] || process.env.XEROLITE_AGENTIC_API_KEY; ... 'X-Agentic-Api-Key': apiKey
Use a revocable, least-privilege key if Xerolite supports it, keep the environment variable private, and point XEROLITE_API_URL only at a trusted Xerolite instance.
Financial positions and performance data could become visible in transcripts or shared logs.
Portfolio queries can return sensitive financial holdings and PnL data that may be printed into the agent conversation, terminal output, or logs.
Response: JSON array of portfolio rows (broker, ticker, position size, market value, PnL fields, etc.)
Only fetch portfolio data when needed, avoid sharing transcripts or verbose logs, and treat outputs from this skill as sensitive financial information.