科技投资日报

WarnAudited by ClawScan on May 10, 2026.

Overview

The report-generation purpose is coherent, but the skill uses local Feishu app secrets and sends a PDF to a hard-coded Feishu user without declaring that credential access or recipient control.

Before installing, verify that you want this skill to use your Feishu app credentials, change or remove the hard-coded Feishu open_id, require confirmation before sending reports, and inspect the referenced PDF converter dependency.

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.

High

#ASI03: Identity and Privilege Abuse

What this means

Installing the skill could allow it to use the user's configured Feishu app credentials to obtain a tenant token and send messages/files.

Why it was flagged

The skill instructs the agent to read Feishu app credentials from the local OpenClaw config, while the supplied metadata declares no primary credential, required env vars, or required config paths.

Skill content

cfg = json.load(open("/root/.openclaw/openclaw.json"))["channels"]["feishu"]; app_id, app_secret = cfg["appId"], cfg["appSecret"]

Recommendation

Declare the Feishu credential/config requirement clearly, use the platform's scoped credential mechanism if available, and require explicit user confirmation before using those credentials.

Medium

#ASI02: Tool Misuse and Exploitation

What this means

A generated PDF report could be sent from the user's Feishu workspace/app to an unintended fixed recipient.

Why it was flagged

The PDF attachment is sent through a raw Feishu API call to a hard-coded open_id rather than a user-selected or clearly confirmed recipient.

Skill content

user_id = "ou_159cbb6a3791ff5a98f3a2a4b38e7d4c" ... requests.post("https://open.feishu.cn/open-apis/im/v1/messages?receive_id_type=open_id", ... json={"receive_id": user_id, "msg_type": "file" ...})

Recommendation

Replace the hard-coded recipient with a user-configured destination and ask for confirmation before sending files or messages.

Low

#ASI04: Agentic Supply Chain Vulnerabilities

What this means

PDF generation depends on external local code whose contents are not part of this review.

Why it was flagged

The workflow relies on another skill's converter script that is not included in this package or declared in the install metadata.

Skill content

bash /root/.openclaw/workspace/skills/md2pdf-weasyprint/scripts/convert-weasyprint.sh /tmp/tech-invest-YYYYMMDD.md /tmp/tech-invest-YYYYMMDD.pdf

Recommendation

Declare the dependency explicitly, include or pin the converter provenance, and let users verify it before execution.

Low

#ASI09: Human-Agent Trust Exploitation

What this means

Users may place too much trust in automated buy/sell ranges, target prices, and stop-loss suggestions.

Why it was flagged

The report template requires concrete trading guidance while also disclaiming that it is investment advice.

Skill content

**投资建议** 建仓区间：XX~XX | 目标价：XX | 止损：XX ... ⚠️ 以上内容仅供参考，不构成投资建议

Recommendation

Make the financial-risk disclaimer prominent and consider requiring the user to opt in before generating actionable trading recommendations.