DevTrendsDeveloper trends
Accent color

hashgraph-online /

hol-guard

Project URL Category Python Project Title Local firewall for AI coding agent tool calls Project Description HOL Guard is an Apache-2.0 local runtime layer for AI coding agents such as Claude Code, Codex and Cursor. It ch…

LanguagePythonDTSecurity & testingDTSoftware developmentDTAgent extensionDTAgent evaluation & safetyDTCLIDTWeb appDTLinuxDTTerminalDTWebDTClaude CodeDTCodexOriginalagent-security

About

Project URL Category Python Project Title Local firewall for AI coding agent tool calls Project Description HOL Guard is an Apache-2.0 local runtime layer for AI coding agents such as Claude Code, Codex and Cursor. It checks shell, secrets, MCP and package-install calls before they run, then allows, blocks or asks. Highlights 1. Checks before execution: policy is applied to the agent's shell commands, secret-file reads, MCP server changes and package installs on your own machine, instead of trusting the model. 2. Local-first: runs on the developer machine; every allow/block/ask decision is kept in a local history you can review. 3. Works across agents: Claude Code, Codex, Cursor, Gemini CLI and other coding agents. 4. Honest scope: open source (Apache-2.0), maker HOL. A scan is not a safety guarantee, and it is not a complete prompt-injection defense. Example Code bash pipx install hol-guard && hol-guard init Screenshots or Demo Videos Docs: Self-recommendation from the maintainers (HOL).

Discovery history

  1. HelloGitHub·用户投稿

    Oct 20, 2026 · 1 observations

Discuss this product