Instructions file
Instructions for jmanhype/vggt-mps, covering claude code guardrails, allowed modifications, restricted modifications, exception: issues labeled ai-implement and pull request requirements.
Instructions file
Claude Code instructions for jmanhype/MCPhoenix, covering claude code guardrails, allowed modifications, restricted modifications, exception: issues labeled ai-implement and pull request requirements.
jmanhype/claude-code-plugin-marketplace
Plugin Claude Code
Production-grade multi-agent systems for trading, automation, and swarm intelligence.
jmanhype/claude-code-plugin-marketplace
Instructions file CodexOpenCode
AGENTS.md instructions for jmanhype/claude-code-plugin-marketplace, covering repository guidelines, project structure & module organization, build, test, and development commands, coding style & naming conventions and testing guidelines.
jmanhype/claude-code-plugin-marketplace
Instructions file
Claude Code instructions for jmanhype/claude-code-plugin-marketplace, covering claude code guardrails, allowed modifications, restricted modifications, exception: issues labeled ai-implement and pull request requirements.
jmanhype/claude-code-plugin-marketplace
Skill Claude CodeCodex
Synthesize Reflector insights into structured delta proposals for playbook updates, following ACE paper's Curator architecture.
jmanhype/claude-code-plugin-marketplace
Skill Claude CodeCodex
Generate Python code to solve AppWorld agent tasks using playbook bullet guidance. Use when the AppWorld executor needs executable Python code for tasks involving Spotify, Venmo, Gmail, Calendar, Contacts, or other AppWorld APIs.
jmanhype/claude-code-plugin-marketplace
Skill Claude CodeCodex
Analyze AppWorld task failures to extract specific API patterns and generate actionable playbook bullets with concrete code examples.
jmanhype/claude-code-plugin-marketplace
Agent Claude Code
You are a specialized AI safety agent that detects backdoors and malicious patterns in code using a DSPy-powered monitor with GEPA-optimized prompts.
jmanhype/claude-code-plugin-marketplace
Agent Claude Code
You are a human-in-the-loop code auditor who reviews code flagged by the safety monitor as potentially malicious.
jmanhype/claude-code-plugin-marketplace
Agent Claude Code
You enforce AI control protocols to ensure safe deployment of untrusted LLMs.
jmanhype/claude-code-plugin-marketplace
Agent Claude Code
You generate natural language feedback for GEPA optimization to improve monitor performance.
jmanhype/claude-code-plugin-marketplace
Agent Claude Code
You are a fine-grained classification specialist that assigns precise suspicion scores (0-100) to code samples.
jmanhype/claude-code-plugin-marketplace
Agent Claude Code
You scan code for known security vulnerabilities, CVEs, and common exploit patterns.
jmanhype/claude-code-plugin-marketplace
Command Claude Code
Review and audit code flagged as suspicious by the monitor.
jmanhype/claude-code-plugin-marketplace
Command Claude Code
Export the optimized monitor as a reusable Claude Skill for integration with other projects.
jmanhype/claude-code-plugin-marketplace
Command Claude Code
Bootstrap fine-tune the monitor after GEPA optimization to further improve performance.
jmanhype/claude-code-plugin-marketplace
Command Claude Code
Load datasets from Control Arena (APPS, ControlTax, AppsBackdoors) for training and evaluation.
jmanhype/claude-code-plugin-marketplace
Command Claude Code
Run GEPA (Genetic-Pareto) prompt optimization to improve monitor performance.
jmanhype/claude-code-plugin-marketplace
Command Claude Code
Scan code for backdoors, malicious patterns, and security vulnerabilities using the DSPy-powered safety monitor.
jmanhype/claude-code-plugin-marketplace
Command Claude Code
Train the DSPy monitor using comparative pairs of attack vs honest code.
jmanhype/claude-code-plugin-marketplace
Command Claude Code
Compute safety scores and evaluate monitor performance across audit budgets.
Agent Claude Code
Specializes in backend API implementation following TDD and contract-driven development. Uses Specmatic for contract testing and validation.
Agent Claude Code
Synchronizes Spec Kit tasks with Beads persistent task memory.