Unified code review engine with 6 specialized agents, adaptive confidence scoring, security blocking, risk classification, and churn detection. Works with or without git.
Deep logic error detector that finds actual bugs — null handling, race conditions, off-by-one errors, type mismatches, infinite loops, and resource leaks. Uses Opus for nuanced reasoning about complex code paths. Works with or without git. <example> Context: Complex algorithm implementation user: "Check this code for bugs" assistant: "I'll launch bug-hunter with Opus to deeply analyze for logic errors." </example>
Post-review code simplification and polish. Runs AFTER other review agents pass. Simplifies code for clarity, reduces unnecessary complexity, and applies project conventions — all while preserving exact behavior. Language-agnostic. <example> Context: Code passed review, needs polish user: "Clean up this code" assistant: "I'll use code-polisher to simplify and polish while preserving behavior." </example>
Silent failure and error handling auditor. Finds empty catch blocks, broad exception catching, missing error propagation, unjustified fallbacks, and inadequate error messages. Inspired by pr-review-toolkit's silent-failure-hunter but language-agnostic. <example> Context: Code with try-catch blocks user: "Check error handling in my changes" assistant: "I'll use error-auditor to hunt for silent failures and inadequate error handling." </example>
Deep security vulnerability scanner using Opus for semantic analysis. Goes beyond pattern matching to understand data flow, trust boundaries, and attack surfaces. Finds injection, auth bypass, data exposure, and OWASP Top 10 vulnerabilities. <example> Context: Code handling user input or authentication user: "Security review this code" assistant: "I'll use security-scanner with Opus for deep security analysis." </example>
General code quality reviewer that checks CLAUDE.md compliance, coding standards, and project conventions. Uses adaptive confidence scoring to minimize false positives. Language-agnostic — auto-detects the project's stack and conventions. <example> Context: User finished implementing a feature user: "Review my recent changes" assistant: "I'll launch the sentinel-reviewer agent to check code quality and CLAUDE.md compliance." </example> <example> Context: Non-git project, files tracked via edit log user: "Check the code I just wrote" assistant: "I'll use sentinel-reviewer to review the files modified this session." </example>
Modifies files
Hook triggers on file write and edit operations
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
By Siddharth Gupta | PretInnov Technologies
Three production-grade plugins that make Claude Code smarter, safer, and capable of building its own extensions.
| Plugin | What It Does | One-Liner |
|---|---|---|
| Clamper | Verifies your code, maps your project DNA, and scaffolds full agent ecosystems | "Clip it, Clamp it." |
| Cortex | Remembers what worked and what failed across ALL your sessions, then uses that knowledge automatically | "Claude Code with a brain." |
| Forge | Creates entire plugins from a simple chat prompt — researches, asks questions, designs, and generates everything | "Build plugins by talking." |
This tells Claude Code "hey, there's a plugin store at this GitHub repo." You only do this once.
Option A — Using the Terminal (CLI):
Open your terminal and type:
claude plugin marketplace add github:PretInnov-Inc/forge-marketplace
Option B — Using the Desktop App:
PretInnov-Inc/forge-marketplaceOption C — From inside a Claude Code chat session:
Just type this as a message:
/plugin marketplace add github:PretInnov-Inc/forge-marketplace
Now you can install any (or all) of the 3 plugins:
# Install all three:
claude plugin install clamper@pretinnov-plugins
claude plugin install cortex@pretinnov-plugins
claude plugin install forge@pretinnov-plugins
Or from inside a Claude Code session:
/plugin install clamper@pretinnov-plugins
/plugin install cortex@pretinnov-plugins
/plugin install forge@pretinnov-plugins
Type /help inside Claude Code — you should see the new commands listed under their plugin namespaces.
/clamper:init # Set up your project with agents, skills, and rules
/clamper:clamp # Check if your recent code changes are safe
/clamper:dna # Analyze your project structure, hot files, fragile zones
/clamper:clamper # View the Clamper dashboard with stats
/cortex:cortex # View what Cortex has learned about you
/cortex:decisions # Track important decisions you've made
/cortex:health # Check system health
/cortex:evolve # Let Cortex adapt its rules based on your feedback
/forge # Create a new plugin (full guided process)
/forge:research <idea> # Research if a plugin idea already exists
/forge:validate <path> # Check if a plugin is correctly built
/forge:dashboard # View your build history
When we release new versions, update like this:
# Update the marketplace catalog first
claude plugin marketplace update pretinnov-plugins
# Then update individual plugins
claude plugin update clamper@pretinnov-plugins
claude plugin update cortex@pretinnov-plugins
claude plugin update forge@pretinnov-plugins
claude plugin uninstall clamper@pretinnov-plugins
claude plugin uninstall cortex@pretinnov-plugins
claude plugin uninstall forge@pretinnov-plugins
# Remove the marketplace entirely
claude plugin marketplace remove pretinnov-plugins
MIT — Use it, modify it, share it. No restrictions.
Siddharth Gupta — PretInnov Technologies
npx claudepluginhub pretinnov-inc/claude-plugin-marketplace --plugin sentinelSelf-learning intelligence layer that auto-extracts learnings, mines patterns, tracks decisions, and continuously optimizes your Claude Code workflow across all projects and sessions.
Meta-plugin that creates production-grade Claude Code plugins through a 4-phase pipeline: internet research, clarifying questions, architecture blueprint, and full code generation. Learns from every build to improve future plugin quality.
Clip it, Clamp it. One-command ecosystem initialization, verification loop, and deep project DNA. /init scaffolds the full cross-platform agent ecosystem for any project.
Automated code review, security scanning, and quality enforcement
Expert code review specialist. Proactively reviews code for quality, security, and maintainability. Use immediately after writing or modifying code.
Personal Claude Code + Codex dev stack: security hooks, AI-first code conventions, /security-review, /repo-map, /stack-check, portable statusline. Designed to complement other skills-based plugins, not replace them.
Live codebase visualization and structural quality gate — 14 health dimensions graded A-F, dependency analysis, and architecture governance via MCP
Universal quality control orchestrator and final authority for any software development project. Dynamically discovers and coordinates with available sub-agents, performs comprehensive multi-dimensional quality assessment, security validation, and deployment readiness verification. Adapts to any project type, programming language, or development framework while maintaining enterprise-grade quality standards. Examples: <example>Context: Code changes ready for review across any project. user: 'Please review this code before commit' assistant: 'I'll use the 1-ceo-quality-control-agent to orchestrate comprehensive quality validation, discover available specialists, and perform final security scanning before approval.' <commentary>Universal quality control requires comprehensive validation across all dimensions regardless of project type.</commentary></example> <example>Context: Multi-agent work completion needing validation. user: 'Several agents completed their tasks, need quality review' assistant: 'Let me engage the 1-ceo-quality-control-agent to coordinate comprehensive validation across all completed work and ensure quality standards.' <commentary>Multi-agent coordination and quality validation applies to any development project.</commentary></example>
Agents for code review, security audits, debugging, and quality assurance