JeongJaeSoon/agent-guard

Agent Guard provides deterministic secret-scanning guardrails for AI coding agents (Claude Code, Codex), Git hooks, GitHub Actions, and a direct CLI. It prevents common accidental secret exposures—reading sensitive files like .env, writing secret-like values, running shell commands that dump credentials, or leaving secrets in the working tree—by using gitleaks for detection and shell scripts for integration. Intended users are developers and teams using AI code assistants or CI who need predictable pre/post-tool checks (it is not a vault, credential rotator, or a replacement for platform secret scanning). Key features: Claude Code and Codex plugins/hooks, native git hook install, GitHub Actions step, checksum-verified installer, CLI commands (scan-working-tree, scan-staged, scan-path, smoke-test, setup, checksum), and a pii-filter that masks or can block PII using a built-in regex provider or external redact endpoints. Requires sh, awk, git, jq, and gitleaks.

AI-related keywords found in this repo's description, topics, or README summary — grouped by category. Each badge links to the corresponding ranking detail page.