---
title: "z"
description: "Make your coding agents automatically self-improve. Enforce your coding standards while your developers keep using whatever IDE, model, or agent they want."
canonical: https://agentpluginsdirectory.com/plugins/z
last-updated: 2026-09-23
---

# z
Make your coding agents automatically self-improve. Enforce your coding standards while your developers keep using whatever IDE, model, or agent they want.
- Slug: z
- Publisher: Zenable
- Repository: https://github.com/Zenable-io/ai-guardrails
- Manifest: plugins/z/plugin.json
- Version: 2.4.3
- License: Apache-2.0
- Category (editorial): code-review
- Skills: 11 (addtests, assessment, debug, doublecheck, feat, guardrails-reviewer, prfeedback, rebase, researchbranch, setup, triage)
- MCP servers: 0
- Stars: 2
- Repository created: 2025-10-13
- Repository last pushed: 2026-09-21
- Publisher type: Organization
- Listing: https://agentpluginsdirectory.com/plugins/z
- Schema: https://agent-plugins.org/schemas/1.0.0/plugin.schema.json

## What z does, in the publisher's words

Zenable enforces your organization's coding standards, the requirements your team decided on, against the code AI agents write, as they write it. Teams adopt AI coding agents at full speed and stay audit-ready, and developers keep whatever IDE or agent they already use.

You describe a standard once as a plain-English requirement. Zenable enforces it with a mix of deterministic static analysis and AI review, and records every finding, a durable account of what was actually checked, not a claim that an agent looked.

This repository is the home for Zenable's developer-side integrations: the agent plugin, which ships as both a Claude Code plugin and a portable Agent Plugins 1.0 package: a GitLab CI/CD component, a pre-commit hook, and drop-in configuration for plenty of different agentic IDE. Your requirements travel with you across all of them.

From the project README, punctuation lightly normalized. Full text: https://raw.githubusercontent.com/Zenable-io/ai-guardrails/HEAD/README.md

## Skills

- addtests: Add tests that cover core concepts and prevent regressions without overfitting or crystallizing the code base. Use when the user asks for tests, more coverage, or a regression test for a bug fix, or invokes `/z:addtests`.
- assessment: Run a guided, end-to-end code assessment: frame the engagement, onboard the codebase's requirements via the setup skill, run `zenable check` plus an AI review pass against those same requirements, merge and verify the findings, and produce a polished HTML report bundle. Use this skill whenever the…
- debug: Expert debugging using the scientific method to identify root causes, forming and falsifying hypotheses against observable evidence. Use when the user reports a bug, a failing test, or unexplained behavior, or invokes `/z:debug`.
- doublecheck: Double check work for quality, test coverage, elegance, hallucinations, and orphaned code left behind by changes on this branch. Use when the user asks you to review or verify your own work before shipping, or invokes `/z:doublecheck`.
- feat: Implement a new feature with research, a PRD, SOLID design, non-overfitted tests, and incremental commits. Use when the user asks to build, add, or implement a feature, or invokes `/z:feat`.
- guardrails-reviewer: Reviews code changes against your Zenable requirements using hybrid LLM-as-judge analysis and Zenable's deterministic guardrails. Automatically invoked when making code changes or at development milestones. Use for security compliance, quality checks, and policy enforcement, or when the user invoke…
- prfeedback: Address every unresolved pull request comment on the current branch's PR by making code changes or replying, then mark each thread as addressed. Use when the user asks to handle PR feedback or review comments from human reviewers, or invokes `/z:prfeedback`.
- rebase: Expert help resolving Git merge conflicts with non-destructive merges that preserve behavior from both sides. Use when the user is merging or rebasing main into a feature branch, hits conflict markers, or invokes `/z:rebase`.
- researchbranch: Research and summarize the intent, current state, rationale, and outstanding work for the current branch, including its pull request status. Use when the user asks what a branch is for, where it stands, or what is left to do, or invokes `/z:researchbranch`.
- setup: Guided Zenable onboarding: discover the standards a codebase already has written down, mine them into candidate requirements, triage them with the user, and persist the survivors as scoped Zenable requirements through the Zenable MCP. Use this skill whenever the user wants to onboard a codebase on…
- triage: Fetch and address unresolved Zenable AI guardrails review comments on the current pull request (GitHub PR) or merge request (GitLab MR). Use this skill whenever the user says "triage", "address PR feedback", "fix review comments", "respond to the guardrails bot", or anything about responding to aut…

Descriptions come from the frontmatter of each SKILL.md, punctuation lightly normalized.