Enforce Coding-Agent Write and Shell Policies with Probity
Use Probity when Claude Code, Codex, or GitHub Copilot CLI should be blocked from unsafe file writes or shell commands before they happen.
npx skills add agentskillexchange/skills --skill enforce-coding-agent-write-and-shell-policies-with-probity
Use Probity when an agent/operator needs deterministic or AI-validated guardrails around coding-agent actions. The operator installs the package in a project, defines scoped rules in probity.config.ts, wires Probity into a supported agent, and lets it inspect file writes and shell commands before they execute. Built-in rules can enforce TDD, block command patterns, require prior test commands, forbid content patterns, or enforce filename casing.
What this skill actually does
Invoke this instead of relying on prompts or normal coding-agent behavior when policy needs to be enforced at the action boundary. It is useful when the team wants failing-test-first discipline, destructive-command blocks, required validation before commits, or custom TypeScript rules that provide the agent a reason and a recovery path.
The scope boundary is pre-action policy enforcement for supported coding agents. This is not just an npm library card or a generic TDD reminder; the skill is about installing and configuring Probity as a guardrail layer that can block unsafe writes and commands in Claude Code, Codex, and GitHub Copilot CLI sessions.