Skill Trust Decision
amazon-screenshot
Legitimate Amazon screenshot automation with hardcoded SMTP credential and unpinned dependencies as primary concerns.
Why this conclusion was reached
2/4 dimensions flagged Block
Declared vs actual capability
1 undeclared or violating capabilities were inferred.
Review
Hidden execution and egress
9 lower-risk artifacts were extracted and still need context.
Block
Attack chain and severe findings
The report includes 0 attack-chain steps and 1 severe findings.
Review
Dependencies and supply chain hygiene
4 dependency or supply-chain issues need attention.
What drove the risk score up
Hardcoded SMTP credential in source +15
SMTP password 'rDs70BwDDg6U9YbO' for [email protected] is embedded in screenshot.js line 54, not declared in SKILL.md, and visible in skill-card.md Known Risks section
Unpinned dependencies +8
package.json uses caret versions (^8.0.7, ^1.59.1) allowing automatic minor version updates
Undeclared shell execution +5
execSync with inline Python script for X11 display detection not mentioned in SKILL.md
Most important evidence
High Credential Theft
Hardcoded SMTP service account credential
SMTP password for [email protected] is embedded as a plain-text string in screenshot.js line 54. While acknowledged in skill-card.md, it is not declared in SKILL.md and creates risk if source is exposed.
scripts/screenshot.js:54 Move SMTP credentials to environment variables or a secrets manager. Rotate the exposed password immediately.
Medium Supply Chain
Unpinned dependency versions
All dependencies use caret (^) versioning allowing automatic updates: nodemailer@^8.0.7, playwright@^1.59.1. Malicious updates could be injected via npm registry.
package.json:3 Pin exact versions ([email protected], [email protected]) to prevent supply chain attacks.
Low Doc Mismatch
Shell execution not declared in SKILL.md
autoDetectDisplay() uses execSync to run an inline Python script querying /tmp/.X11-unix/ for X11 sockets. This shell execution is not mentioned in SKILL.md documentation.
scripts/screenshot.js:35 Document shell execution capability in SKILL.md or refactor to use Node.js native APIs.
Low Sensitive Access
Persistent Chrome browser profile storage
The script uses a persistent Chrome profile (chrome_profile/) to maintain login sessions and cookies, which persists sensitive authentication data.
scripts/screenshot.js:65 Use ephemeral browser contexts without persistent storage for better security isolation.
Declared capability vs actual capability
Filesystem Pass
Declared WRITE
→ Inferred WRITE
Creates screenshots and ZIP files in outputs/ directory Network Pass
Declared READ
→ Inferred READ
Fetches Amazon pages via Playwright browser automation Shell Block
Declared NONE
→ Inferred WRITE
scripts/screenshot.js:35 - execSync runs Python for X11 detection Browser Pass
Declared READ
→ Inferred READ
Playwright with persistent Chrome profile for Amazon automation Suspicious artifacts and egress
Medium External URL
https://www.amazon.com/https://www.amazon.com/ chrome_profile/Default/Service Worker/CacheStorage/2a1c18365e83cb6e1347464058faa3d44e45a31c/index.txt:11
Medium External URL
https://www.patreon.com/feross scripts/package-lock.json:84
Medium External URL
https://feross.org/support scripts/package-lock.json:88
Medium External URL
https://www.amazon.com/dp/$ scripts/screenshot.js:65
Medium External URL
https://www.amazon.com scripts/screenshot.js:135
Medium External URL
https://clawhub.ai/user/328786889 skill-card.md:7
Medium External URL
https://clawhub.ai/328786889/amazon-screenshot skill-card.md:29
Dependencies and supply chain
| Package | Version | Source | Known vuln | Notes |
|---|---|---|---|---|
| nodemailer | ^8.0.7 | npm | No | Version not pinned |
| playwright | ^1.59.1 | npm | No | Version not pinned |
| playwright-extra | ^4.3.6 | npm | No | Version not pinned |
| playwright-extra-plugin-stealth | ^0.0.1 | npm | No | Version not pinned |
File composition
11 files · 1480 lines
JSON 7 files · 979 linesJavaScript 1 files · 373 linesMarkdown 2 files · 118 linesText 1 files · 10 lines
Files of concern · 6
scripts/package-lock.json https://www.patreon.com/feross · https://feross.org/support
scripts/screenshot.js Hardcoded SMTP service account credential · Shell execution not declared in SKILL.md · Persistent Chrome browser profile storage · https://www.amazon.com/dp/$ · https://www.amazon.com
SKILL.md skill-card.md https://clawhub.ai/user/328786889 · https://clawhub.ai/328786889/amazon-screenshot
chrome_profile/Default/Service Worker/CacheStorage/2a1c18365e83cb6e1347464058faa3d44e45a31c/index.txt https://www.amazon.com/https://www.amazon.com/
package.json Unpinned dependency versions
Other files · package-lock.json · package.json · _meta.json · metadata.json · metadata.json
Security positives
Legitimate business functionality with clear user consent model (requires ASINs and email from user)
Intentional email domain restriction (@campsnail.com only) limits misuse scope
Automatic cleanup of temporary screenshots and ZIP files after email delivery
Lock file mechanism prevents concurrent execution abuse
No evidence of data exfiltration or unauthorized credential harvesting from host environment
Stealth automation and persistent login states are declared in SKILL.md under '反爬注入' and '持久化登录状态'