Block Risky Coding-Agent Commands with CC Safety Net
Use CC Safety Net when coding-agent CLIs need pre-execution hooks that block destructive commands, secret access, and unsafe file operations before tools run.
npx skills add agentskillexchange/skills --skill block-risky-coding-agent-commands-with-cc-safety-net
Use CC Safety Net when an operator needs hard pre-execution controls around coding-agent CLIs. The operator installs CC Safety Net into one or more supported hosts, configures the policy mode and any custom rulebook, then uses doctor, explain, logs, or the local GUI to verify and audit decisions before relying on the integration.
What this skill actually does
Invoke this instead of using the coding-agent product normally when instructions and review discipline are not enough: destructive Git commands, recursive deletes, secret-file reads, credential-store access, and unsafe shell wrappers should be intercepted before the agent tool call executes.
The scope boundary is pre-tool command and file-access safety for coding-agent CLIs. It is not a general sandbox, endpoint security platform, or policy framework; it installs a concrete hook layer with documented protections, diagnostics, and audit logs across supported agent hosts.
Inputs and prerequisites: Node.js/npx or npm; CC Safety Net CLI; one supported coding-agent host such as Claude Code, Codex, Gemini CLI, GitHub Copilot CLI, Kimi Code, OpenCode, Pi, Cursor, Amp Code, Hermes Agent, Antigravity CLI, or OpenClaw..
Setup notes: Run npx -y cc-safety-net@latest install and select the installed coding-agent CLIs to protect. For a global CLI alias, run npm install -g cc-safety-net. After installation, run npx cc-safety-net doctor to verify the hook integration.
Source and verification boundary: use https://ccsafetynet.com/docs as the canonical reference before running the workflow; keep commands, API calls, CLI usage, and generated outputs reviewable against that upstream source.
Framework fit: publish this as a Multi-Framework workflow only when the operator can invoke the documented toolchain directly, rather than treating the upstream project as a generic product listing.