Project Codeguard Rules
- 418 repo stars
- Updated January 29, 2026
- project-codeguard/rules
Project CodeGuard is a Claude Code plugin that embeds secure-by-default security rules, agent translators, and validators into AI coding and review workflows.
About
Project CodeGuard is a Claude Code plugin packaging a security framework and ruleset meant to sit inside AI coding workflows—not replace your scanner, but steer models toward secure-by-default patterns during generation and review. developers who ship fast with Claude Code install it when a single prompt slip can introduce injection, secrets leakage, or weak auth and they want consistent guardrails across sessions. The project ships core rules, agent-specific translators so instructions map to Claude and other hosts, and validators to check compliance with those rules. It pairs well with review habits: generate with constraints, then validate before merge. It does not deploy infrastructure or fix CVEs automatically; it standardizes how your agent writes and critiques code. Catalog metadata lists one plugin and keywords around agents, codeguard, coding, comprehensive, rules, and security.
- AI model-agnostic security framework for coding workflows
- Core security rules with translators for popular coding agents
- Validators to test whether generated or reviewed code meets rules
- Covers both AI code generation and AI-assisted review paths
- Community catalog bundle with comprehensive rules and agents keywords
Project Codeguard Rules by the numbers
- Data as of Jul 28, 2026 (Skillselion catalog sync)
/plugin install project-codeguard-rules@project-codeguard/rulesAdd your badge
Show developers this plugin is listed on Skillselion. Paste this into your README.
| repo stars | ★ 418 |
|---|---|
| Last updated | January 29, 2026 |
| Repository | project-codeguard/rules ↗ |
What it does
Embed secure-by-default rules into AI-assisted coding and reviews so Claude-generated code is checked against a shared security framework.
Who is it for?
Best when you're relying on Claude Code for most implementation and want agent-agnostic security rules and compliance checks without building custom prompts from scratch.
Skip if: Skip if you have no AI in the loop and only need traditional SAST in CI, or developers and want penetration testing services instead of coding guardrails.
What you get
Your Claude Code sessions follow a shared CodeGuard rules framework with validators so generated and reviewed code aligns to secure-by-default expectations before you ship.
- Agent-ready security rules and translators applied to your coding workflow
- Repeatable validator runs against AI-generated or reviewed code
- Shared secure-by-default baseline across generation and review sessions
By the numbers
- pluginCount: 1 in catalog metadata
- Repository: project-codeguard/rules
- Catalog keywords: agents, codeguard, coding, comprehensive, rules, security
Recommended Plugins
1mangesh1 Hipaa GuardianAI agent skill for HIPAA compliance - PHI/PII detection, code scanning, audit reports, and remediation guidance6
Abossenbroek Abossenbroek Claude PluginsCustom Claude Code plugins by abossenbroek2
Agentsecops SecopsagentkitSecurity operations toolkit for AI coding agents.181
Agent Sh EnhancePlugin structure and tool use analyzer - validates plugin.json, MCP tools, and security patterns3
Altic Dev EnvaultLet agents safely manage env variables6
Auth0 Agent SkillsAuth0 Agent Skills40How it compares
Security rules and validators for AI coding workflows, not OpenAPI SDK generation or agent long-term memory.
FAQ
Who is Project Codeguard Rules for?
Developers and small teams using AI coding agents who need consistent secure-by-default rules during generation and review.
When should I use Project Codeguard Rules?
Before you merge or ship features written heavily with Claude Code, or when you are standardizing security expectations across multiple agent tools.
How do I add Project Codeguard Rules to my agent?
Install the project-codeguard/rules Claude Code plugin from the catalog, apply the bundled rules and agent translators in your Claude Code configuration, and run the packaged validators on generated or reviewed code per the repo README.