The captcha-solve-api PyPI package is a malicious package masquerading as a captcha-solving API client but primarily delivers a telemetry-based backdoor. It installs two large modules (_telemetry_init and _telemetry_transport) and a .pth file that auto-executes on Python startup, independent of the advertised API usage. The telemetry modules collect host identifiers and send them to a remote server whose address is dynamically reconstructed via DNS TXT queries to public resolvers, enabling endpoint rotation and evasion of allowlists. The package also attempts to download and execute malicious binaries from Cloudflare Workers or via DNS TXT records from a set of domains, then deletes the binaries after execution. The install command is overridden to run malicious code during installation, ensuring persistence and stealth. Versions 0.0.1 contain disarmed code, with active malicious functionality in later versions 8.5.3 and 8.5.4.

This malicious package compromises systems by installing persistent backdoor telemetry modules that exfiltrate system information. It enables remote code execution by downloading and running malicious binaries. The use of DNS-based command and control and dynamic endpoint resolution allows evasion of network defenses and complicates detection and blocking. The persistence mechanism ensures the malicious code runs on every Python interpreter start, potentially affecting any Python environment using this package. The package abuses PyPI distribution to spread malware under the guise of a legitimate captcha-solving client.

No official patch or remediation is currently available. Users should immediately uninstall the captcha-solve-api package if installed and avoid using versions 0.0.1, 8.5.3, and 8.5.4. Since the malicious code persists via a .pth file and global site modules, manual removal of these artifacts from site-packages is necessary to fully clean affected environments. Monitor for suspicious DNS TXT queries and network connections to unknown endpoints. Consider restricting DNS queries to trusted resolvers and blocking known malicious domains. Check vendor advisories or PyPI for updates or removal of the malicious package.