insane-search is a public-page reader for Claude Code, enabling access to blocked websites without API keys or proxy setup.
Source: Official README and linked documentation View on GitHub →insane-search is gaining attention for its ability to bypass website blocks and CAPTCHAs, offering a unique solution for accessing public content without authentication. Its adaptive scheduler and TLS impersonation is innovative technical choices.
Source: Official README and linked documentationPhase 0→3 adaptive scheduler that tries various routes to access blocked content, including public APIs, syndication gateways, and TLS impersonation.
Source: Official README and linked documentationTLS impersonation to mimic a full browser identity, enhancing the likelihood of accessing blocked content.
Source: Official README and linked documentationAuto-installs necessary tools on first use without requiring API keys or signups.
Source: Official README and linked documentationThe architecture is modular, with a clear separation of concerns. It includes an adaptive scheduler, a TLS impersonation module, and a headless browser for accessing content. Data flows through various phases, each attempting different methods to bypass blocks.
Source: Official repository tree and dependency filescurl_cffiyt-dlpinsane-search is suitable for users who need to access blocked content on various platforms, such as Reddit, YouTube, and Hacker News. It is useful for researchers, content curators, and anyone who encounters website blocks while working with Claude Code.
Source: Official README and linked documentationv0.16.3 was published on 2026-09-08. It added --json-content for returning article text and trace in one call while keeping --json content-free, made the PDF parser lazy, and added URL-credential masking, single-fetch, and PDF fallback tests.
Source: github.com/fivetaku/insane-search/releases/tag/v0.16.3status: docs-only | checked_at: 2026-09-23 | environment: The official v0.16.3 tag, README, CHANGELOG, plugin manifest, and SKILL.md were reviewed. The plugin was not installed in Claude Code and no third-party site was accessed. | limitations: No local installation or runtime verification was performed. Live site success, CAPTCHA behavior, browser fingerprinting, dependency installation, resource use, privacy impact, and regional legal or site-policy differences were not tested. | observed_output: The v0.16.3 tag matches the manifest. Source documentation confirms that login and paywall flows stop, browser dependencies install under the user directory, and proxy output is treated as untrusted content.
Source: github.com/fivetaku/insane-search/releases/tag/v0.16.3Compared with Playwright, insane-search tries public APIs, feeds, and lightweight requests before using a browser fallback, and adds trace output and an untrusted-content boundary. Playwright is a cross-browser automation and testing framework with fuller page control and assertions. Evaluate insane-search for public article retrieval; use Playwright for repeatable interaction tests, locators, and multi-browser coverage.
Source: github.com/fivetaku/insane-search (1) · github.com/microsoft/playwright (2)insane-search is a tool for accessing blocked content on various platforms. It is particularly useful for users who frequently encounter website blocks and require a solution. Its modular architecture and innovative features make it a project worth watching for those in need of reliable content access.
The v0.16.3 tag matches the manifest. Source documentation confirms that login and paywall flows stop, browser dependencies install under the user directory, and proxy output is treated as untrusted content. No local installation or runtime verification was performed.
The review environment was: The official v0.16.3 tag, README, CHANGELOG, plugin manifest, and SKILL.md were reviewed. The plugin was not installed in Claude Code and no third-party site was accessed. insane-search behavior outside that scope still needs a local check.