v1.2.0

Wyckoff Screen

ReviewClawScan verdict for this skill. Analyzed May 1, 2026, 8:12 AM.

Analysis

The stock-screening behavior is mostly purpose-aligned, but the skill embeds an undeclared data-service token and routes Tushare requests to an undisclosed plain-HTTP IP endpoint.

GuidanceReview this skill before installing or running it. The core stock-screening logic appears aligned with its description, but you should remove or replace the embedded Tushare token, avoid the custom HTTP endpoint unless you trust it, document all provider dependencies, and treat all stock recommendations as non-guaranteed research rather than trading instructions.

Findings (4)

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.

Abnormal behavior control

Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.

Agentic Supply Chain Vulnerabilities

SeverityLowConfidenceHighStatusNote

scripts/screen_data.py

三源优先级: tushare → baostock → efinance

The runtime fetch path includes Tushare and Baostock before Efinance, while the SKILL.md dependency section only highlights efinance, akshare, and scipy.

User impactUsers may not realize the skill can depend on additional data providers and packages beyond those described in the main instructions.

RecommendationDocument all external providers and Python packages, pin versions where practical, and explain the fallback order.

Human-Agent Trust Exploitation

SeverityLowConfidenceHighStatusNote

SKILL.md

S（≥75分）：强烈推荐; A（60~74分）：满足买入条件

The skill presents ranked stock candidates with strong buy-oriented labels, even though it does not execute trades.

User impactA user could over-trust the generated candidates as investment advice rather than as an analytical screen.

RecommendationTreat outputs as research signals only, verify them independently, and avoid making trades solely from the skill's ranking.

Permission boundary

Checks whether tool use, credentials, dependencies, identity, account access, or inter-agent boundaries are broader than the stated purpose.

Identity and Privilege Abuse

SeverityMediumConfidenceHighStatusConcern

scripts/ts_data.py

TOKEN = '676658a3dc1be0f7...'; _http_url = 'http://140.143.209.128:5000'

A provider credential is embedded directly in the skill and paired with a custom endpoint, despite the registry declaring no required credentials or environment variables.

User impactUsers would be relying on an embedded credential they cannot scope, rotate, or verify, and the data access path could fail or be controlled outside the user's account.

RecommendationRemove the hardcoded token, declare any required credential, load it from a user-controlled environment variable, and document the exact provider account and permissions needed.

Sensitive data protection

Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.

Insecure Inter-Agent Communication

SeverityMediumConfidenceHighStatusConcern

scripts/ts_data.py

pro._DataApi__http_url = _http_url

The script overrides Tushare's API URL to a plain-HTTP IP address, making the provider identity and data boundary unclear and exposing the embedded token and query traffic to that service.

User impactThe endpoint could observe requests, return manipulated market data, or expose the embedded token over an unencrypted connection, which may affect screening results.

RecommendationUse the official documented HTTPS provider endpoint, or make any proxy endpoint explicit, configurable, and reviewed before use.