Collection of plugins for gptme.
ACE (Agentic Context Engineering) context optimization plugin with hybrid retrieval, semantic matching, and context curation.
Use for: Optimizing context for better agent performance, semantic search, context budget management.
Attention tracking and routing plugin combining history tracking and HOT/WARM/COLD tier management. Implements dynamic context management for both meta-learning and token optimization.
Use for: Reducing token usage by dynamically loading context based on relevance, analyzing context patterns, improving keyword matching.
Full Claude Code integration plugin. Spawn Claude Code subagents from within gptme for analysis, Q&A, bug fixes, and implementation.
Use for: Security audits, code reviews, codebase Q&A, bug fixes, feature implementation.
Multi-model consensus decision-making system that orchestrates multiple LLMs to provide diverse perspectives and synthesize consensus responses.
Use for: Important decisions, architectural choices, code review from multiple perspectives, model comparison.
Example implementations of gptme hooks for customizing agent behavior.
Use for: Learning how to create custom hooks, template for new hook development.
Work persistence plugin for session continuity. Saves and restores work state across sessions.
Use for: Resuming work after interruptions, maintaining context across sessions.
Multi-provider image generation supporting Google Gemini (Imagen), OpenAI DALL-E, and more with a unified interface.
Use for: Creating diagrams, UI mockups, presentation graphics, visual prototyping.
Language Server Protocol integration for enhanced code intelligence.
Use for: Code completion, diagnostics, and navigation within gptme.
Iterative execution plugin implementing Ralph loops for complex multi-step tasks with automatic iteration and convergence.
Use for: Complex tasks requiring multiple iterations, self-correcting workflows, tasks that need refinement cycles.
Enhanced search capabilities with Warp-style filtering and presentation.
Use for: Fast, intuitive code search with visual highlighting.
Wrapped tool definitions for safer, constrained tool execution.
Use for: Creating sandboxed tool environments with restricted capabilities.
Add to your gptme.toml:
[plugins]
paths = ["path/to/gptme-contrib/plugins"]
enabled = ["gptme_attention_tracker"]All plugins in this collection follow the naming convention:
- Directory:
gptme-<name>/(with hyphen) - Package name:
gptme-<name>(with hyphen, in pyproject.toml) - Python module:
gptme_<name>(with underscore, for imports)
This ensures unique, valid package names and clear identification as gptme plugins.
Once configured, plugins are automatically loaded.