Infisical is the open-source platform for securing secrets, certificates, access, and AI Agents.
AI-safe .env files: Schemas for agents, Secrets for humans.
Make your own running home page
Open-source sandboxed agent harness for teams. Giving every employee a secured personal agent.
Pluggable linting tool to prevent committing credential.
Secrets management for teams and AI agents.
Local gateway for Claude Code, Codex and other AI clients on macOS, Windows and Linux: switch upstreams without touching clients, keep API keys from relays, cut off malicious tool calls, trace every request.
Open source code review that runs in your coding agent before you push. Scanners plus a separate AI reviewer, on your own login, no key.
AI gets the context. Not your private data. Local-first privacy proxy for browser chat, AI APIs, and coding agents.
Ray Amjad's Claude Code mods: panes, bands, prompt guards, secret handling and memory tools. One marketplace, each mod installs on its own.
Framework-agnostic log sanitizer for browsers and Node.js. Redacts secrets and infrastructure identifiers with stable HMAC tokens so event correlation still works. Zero runtime dependencies, extensible rules, CLI included.
One beautiful HTML test report for Playwright, Cypress and WebdriverIO, with Cucumber BDD. Shows what broke, who owns it and whether it is new. Plain-language errors, secrets masked, no server, no API keys.
Scan your Pi sessions for secrets using Betterleaks
Free client-side password and key generator for secure passwords, API keys, UUIDs, JWT secrets, and encryption keys. Powered by the Web Crypto API.