Exposed secret literal
Critical
- Finding
- File appears to expose a hardcoded API secret or token.
captcha-login-assistant
Security checks across static analysis, malware telemetry, and agentic risk
Overview
This instruction-only skill is designed to automate CAPTCHA-protected logins with browser scripting and passwords, which can bypass security controls on arbitrary websites.
Review carefully before use. Only use this on websites or internal systems where you have explicit permission to automate CAPTCHA-protected logins, manually verify the target URL, approve each credential submission, and delete any saved screenshots afterward.
Static analysis
VirusTotal
VirusTotal findings are pending for this skill version.
Risk analysis
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.
What this means
The agent could automate logins in ways a website is explicitly trying to prevent, including submitting credentials and CAPTCHA attempts without enough human oversight.
Why it was flagged
This directs the agent to use browser tooling to solve a CAPTCHA and submit a login, which can bypass an anti-automation security control on arbitrary sites.
Skill content
employs AI vision to recognize CAPTCHA codes, automatically fills in account credentials and CAPTCHA, and completes the login operation
Recommendation
Use only on systems you own or are explicitly authorized to automate, and require manual confirmation before submitting credentials or CAPTCHA responses.
What this means
Passwords or privileged accounts could be submitted to the wrong site or used beyond the user's intended authorization boundary.
Why it was flagged
The workflow depends on account credentials and an arbitrary target login page, but it does not define authorization limits, target restrictions, or how credential use is controlled.
Skill content
Target login page URL; Valid account credentials (username and password)
Recommendation
Provide credentials only for trusted, authorized targets, verify the URL manually, and avoid using privileged accounts unless absolutely necessary.
What this means
A user may be led to approve repeated automated login attempts that violate site rules, trigger account lockouts, or bypass security protections.
Why it was flagged
The wording frames avoidance of login security mechanisms as a normal best practice, which may encourage unsafe trust in automation that evades protective controls.
Skill content
Be aware of system login rate limits to avoid triggering security mechanisms
Recommendation
Do not use this skill to evade rate limits, CAPTCHA protections, or other anti-abuse controls; prefer manual login or administrator-approved automation.
What this means
Sensitive information may remain in screenshot files or model context after the login attempt.
Why it was flagged
The workflow stores login-page screenshots for AI vision; such screenshots can contain sensitive page, account, or session information even though this is disclosed and purpose-aligned.
Skill content
mcp_chrome-devtools-mcp_take_screenshot({ "filePath": "path/to/login_screenshot.png" })Recommendation
Save screenshots only in safe locations, delete them after use, and avoid capturing pages that reveal unnecessary private data.