Endor Labs vs KeyHog

Endor Labs is an application security platform spanning code, secrets, dependencies, container images and coding agents. AURI for Developers helps scan for vulnerabilities, detect secrets and block malicious dependencies in AI coding workflows. Endor Code provides AI SAST and secrets detection; Endor Open Source adds reachability-based SCA, malicious-package detection, AI model governance, and SBOM and VEX generation. The platform can inventory coding agents, models, MCP servers and skills, then enforce policies on agent actions. Integrations include GitHub, GitLab, Bitbucket, CircleCI, Jenkins, Jira, Slack, Vanta, Cursor, Claude, Gemini and GitHub Copilot. Scans can run through cloud apps, inside CI/CD runners or with Endor Outpost for scheduled monitoring and on-premises deployment. The company says cloud scanning briefly clones code into a container that is destroyed after scanning, while CI/CD scanning keeps code in the runner. The free Developer tier supports local scans and read-only vulnerability data, but has no UI, policies or scan history. Paid plans are seat-based and require contacting sales for pricing.

KeyHog is free, open-source Rust software for finding and verifying exposed API keys, tokens, passwords, and other credentials. It scans source repositories and Git history, archives, Docker images, URLs and HAR captures, collaboration content, VCS inventories, cloud-storage inventories, browser assets, and running systems. The project lists 934 service-specific detectors, decode-before-scanning, contextual evidence and suppression, and optional live provider verification. GPU execution can use CUDA, native Metal, or portable WGPU, with calibration selecting a parity-backed option. Its GitHub Action scans checked-out trees, fails on high or critical findings, uploads SARIF to Code Scanning, and retains a workflow artifact. Documented CI integrations include GitLab, CircleCI, Jenkins, Buildkite, and shell jobs. A Unix daemon checks changed repository blobs before commits; Windows uses in-process scanning. KeyHog runs on Linux, macOS, Windows, and self-hosted deployments, and requires Rust 1.89 or newer. Code and detector TOMLs are licensed MIT OR Apache-2.0. The plan costs nothing.