Documentation Added: - skills/chiron-core/references/work-para-structure.md: Complete work PARA guide - skills/chiron-core/references/work-quickstart.md: User quick start guide What Was Completed: 1. ✅ Created outline skill with full MCP integration 2. ✅ Enhanced basecamp skill with project mapping 3. ✅ Enhanced daily-routines with work context 4. ✅ Created Teams transcript workflow guide 5. ✅ Set up PARA work structure (10 projects + 5 areas) 6. ✅ Created comprehensive documentation Integration Ready: - Basecamp ↔ Obsidian: Project mapping and task sync - Outline ↔ Obsidian: Wiki search, export, AI queries - Teams → Obsidian → Basecamp: Transcript processing workflow - All integrated into daily/weekly routines PARA Work Structure: - 01-projects/work/: 10 project folders (placeholders ready for customization) - 02-areas/work/: 5 ongoing areas - 03-resources/work/wiki-mirror/: Ready for Outline exports - 04-archive/work/: Ready for completed work Next Steps for User: 1. Customize project names with actual Basecamp projects 2. Configure Outline MCP with your instance 3. Test Basecamp connection 4. Process first Teams transcript using workflow 5. Add n8n workflows when ready (automate Basecamp/Outline sync) Note: All work knowledge stored in Obsidian (tool-agnostic). Jobs easily portable: archive work/, update tool configs, create new projects.
Opencode Agent Skills & Configurations
Central repository for Opencode Agent Skills, AI agent configurations, custom commands, and AI-assisted workflows. This is an extensible framework for building productivity systems, automations, knowledge management, and specialized AI capabilities.
🎯 What This Repository Provides
This repository serves as a personal AI operating system - a collection of skills, agents, and configurations that extend Opencode's capabilities for:
- Productivity & Task Management - PARA methodology, GTD workflows, project tracking
- Knowledge Management - Note-taking, research workflows, information organization
- Communications - Email management, meeting scheduling, follow-up tracking
- AI Development - Tools for creating new skills and agent configurations
- Memory & Context - Persistent memory systems, conversation analysis
- Custom Workflows - Domain-specific automation and specialized agents
📂 Repository Structure
.
├── agent/ # Agent definitions (agents.json)
├── prompts/ # Agent system prompts (chiron.txt, chiron-forge.txt)
├── context/ # User profiles and preferences
│ └── profile.md # Work style, PARA areas, preferences
├── command/ # Custom command definitions
│ └── reflection.md
├── skill/ # Opencode Agent Skills (11+ skills)
│ ├── task-management/ # PARA-based productivity
│ ├── skill-creator/ # Meta-skill for creating skills
│ ├── reflection/ # Conversation analysis
│ ├── communications/ # Email & messaging
│ ├── calendar-scheduling/ # Time management
│ ├── mem0-memory/ # Persistent memory
│ ├── research/ # Investigation workflows
│ ├── knowledge-management/ # Note capture & organization
│ ├── basecamp/ # Basecamp project management
│ ├── brainstorming/ # Ideation & strategic thinking
│ └── plan-writing/ # Project planning templates
├── scripts/ # Repository utility scripts
│ └── test-skill.sh # Test skills without deploying
├── .beads/ # Issue tracking database
├── AGENTS.md # Developer documentation
└── README.md # This file
🚀 Getting Started
Prerequisites
- Opencode - AI coding assistant (opencode.dev)
- Nix (optional) - For declarative deployment via home-manager
- Python 3 - For skill validation and creation scripts
- bd (beads) (optional) - For issue tracking
Installation
Option 1: Nix Flake (Recommended)
This repository is consumed as a non-flake input by your NixOS configuration:
# In your flake.nix
inputs.agents = {
url = "git+https://code.m3ta.dev/m3tam3re/AGENTS";
flake = false; # Pure files, not a Nix flake
};
# In your home-manager module (e.g., opencode.nix)
xdg.configFile = {
"opencode/skill".source = "${inputs.agents}/skill";
"opencode/context".source = "${inputs.agents}/context";
"opencode/command".source = "${inputs.agents}/command";
"opencode/prompts".source = "${inputs.agents}/prompts";
};
# Agent config is embedded into config.json, not deployed as files
programs.opencode.settings.agent = builtins.fromJSON
(builtins.readFile "${inputs.agents}/agent/agents.json");
Rebuild your system:
home-manager switch
Note: The agent/ directory is NOT deployed as files. Instead, agents.json is read at Nix evaluation time and embedded into the opencode config.json.
Option 2: Manual Installation
Clone and symlink:
# Clone repository
git clone https://github.com/yourusername/AGENTS.git ~/AGENTS
# Create symlink to Opencode config directory
ln -s ~/AGENTS ~/.config/opencode
Verify Installation
Check that Opencode can see your skills:
# Skills should be available at ~/.config/opencode/skill/
ls ~/.config/opencode/skill/
🎨 Creating Your First Skill
Skills are modular packages that extend Opencode with specialized knowledge and workflows.
1. Initialize a New Skill
python3 skill/skill-creator/scripts/init_skill.py my-skill-name --path skill/
This creates:
skill/my-skill-name/SKILL.md- Main skill documentationskill/my-skill-name/scripts/- Executable code (optional)skill/my-skill-name/references/- Reference documentation (optional)skill/my-skill-name/assets/- Templates and files (optional)
2. Edit the Skill
Open skill/my-skill-name/SKILL.md and customize:
---
name: my-skill-name
description: What it does and when to use it. Include trigger keywords.
compatibility: opencode
---
# My Skill Name
## Overview
[Your skill instructions for Opencode]
3. Validate the Skill
python3 skill/skill-creator/scripts/quick_validate.py skill/my-skill-name
4. Test the Skill
Test your skill without deploying via home-manager:
# Use the test script to validate and list skills
./scripts/test-skill.sh my-skill-name # Validate specific skill
./scripts/test-skill.sh --list # List all dev skills
./scripts/test-skill.sh --run # Launch opencode with dev skills
The test script creates a temporary config directory with symlinks to this repo's skills, allowing you to test changes before committing.
📚 Available Skills
| Skill | Purpose | Status |
|---|---|---|
| task-management | PARA-based productivity with Obsidian Tasks integration | ✅ Active |
| skill-creator | Guide for creating new Opencode skills | ✅ Active |
| reflection | Conversation analysis and skill improvement | ✅ Active |
| communications | Email drafts, follow-ups, message management | ✅ Active |
| calendar-scheduling | Time blocking, meeting management | ✅ Active |
| mem0-memory | Persistent memory storage and retrieval | ✅ Active |
| research | Investigation workflows, source management | ✅ Active |
| knowledge-management | Note capture, knowledge organization | ✅ Active |
| basecamp | Basecamp project & todo management via MCP | ✅ Active |
| brainstorming | General-purpose ideation with Obsidian save | ✅ Active |
| plan-writing | Project plans with templates (kickoff, tasks, risks) | ✅ Active |
🤖 AI Agents
Chiron - Personal Assistant
Configuration: agent/agents.json + prompts/chiron.txt
Chiron is a personal AI assistant focused on productivity and task management. Named after the wise centaur from Greek mythology, Chiron provides:
- Task and project management guidance
- Daily and weekly review workflows
- Skill routing based on user intent
- Integration with productivity tools (Obsidian, ntfy, n8n)
Modes:
- Chiron (Plan Mode) - Read-only analysis and planning (
prompts/chiron.txt) - Chiron-Forge (Worker Mode) - Full write access with safety prompts (
prompts/chiron-forge.txt)
Triggers: Personal productivity requests, task management, reviews, planning
🛠️ Development Workflow
Issue Tracking with Beads
This project uses beads for AI-native issue tracking:
bd ready # Find available work
bd create "title" # Create new issue
bd update <id> --status in_progress
bd close <id> # Complete work
bd sync # Sync with git
Quality Gates
Before committing:
- Validate skills:
./scripts/test-skill.sh --validateorpython3 skill/skill-creator/scripts/quick_validate.py skill/<name> - Test locally:
./scripts/test-skill.sh --runto launch opencode with dev skills - Check formatting: Ensure YAML frontmatter is valid
- Update docs: Keep README and AGENTS.md in sync
Session Completion
When ending a work session:
- File beads issues for remaining work
- Run quality gates
- Update issue status
- Push to remote (mandatory):
git pull --rebase bd sync git push - Verify changes are pushed
See AGENTS.md for complete developer documentation.
🎓 Learning Resources
Essential Documentation
- AGENTS.md - Complete developer guide for AI agents
- skill/skill-creator/SKILL.md - Comprehensive skill creation guide
- skill/skill-creator/references/workflows.md - Workflow pattern library
- skill/skill-creator/references/output-patterns.md - Output formatting patterns
Skill Design Principles
- Concise is key - Context window is a shared resource
- Progressive disclosure - Load information as needed
- Appropriate freedom - Match specificity to task fragility
- No extraneous files - Keep skills focused and minimal
Example Skills to Study
- task-management/ - Full implementation with Obsidian Tasks integration
- skill-creator/ - Meta-skill with bundled resources
- reflection/ - Conversation analysis with rating system
- basecamp/ - MCP server integration with multiple tool categories
- brainstorming/ - Framework-based ideation with Obsidian markdown save
- plan-writing/ - Template-driven document generation
🔧 Customization
Modify Agent Behavior
Edit agent/agents.json for agent definitions and prompts/*.txt for system prompts:
agent/agents.json- Agent names, models, permissionsprompts/chiron.txt- Chiron (Plan Mode) system promptprompts/chiron-forge.txt- Chiron-Forge (Worker Mode) system prompt
Note: Agent changes require home-manager switch to take effect (config is embedded, not symlinked).
Update User Context
Edit context/profile.md to configure:
- Work style preferences
- PARA areas and projects
- Communication preferences
- Integration status
Add Custom Commands
Create new command definitions in command/ directory following the pattern in command/reflection.md.
🌟 Use Cases
Personal Productivity
Use the PARA methodology with Obsidian Tasks integration:
- Capture tasks and notes quickly
- Run daily/weekly reviews
- Prioritize work based on impact
- Batch similar tasks for efficiency
Knowledge Management
Build a personal knowledge base:
- Capture research findings
- Organize notes and references
- Link related concepts
- Retrieve information on demand
AI-Assisted Development
Extend Opencode for specialized domains:
- Create company-specific skills (finance, legal, engineering)
- Integrate with APIs and databases
- Build custom automation workflows
- Deploy via Nix for reproducibility
Team Collaboration
Share skills and agents across teams:
- Document company processes as skills
- Create shared knowledge bases
- Standardize communication templates
- Build domain expertise libraries
🤝 Contributing
This is a personal repository, but the patterns and structure are designed to be reusable:
- Fork this repository
- Customize for your own use case
- Share interesting skills and patterns
- Learn from the skill-creator documentation
📝 License
This repository contains personal configurations and skills. Feel free to use the patterns and structure as inspiration for your own setup.
🔗 Links
- Opencode - AI coding assistant
- Beads - AI-native issue tracking
- PARA Method - Productivity methodology
- Obsidian - Knowledge management platform
🙋 Questions?
- Check
AGENTS.mdfor detailed developer documentation - Review existing skills in
skill/for examples - See
skill/skill-creator/SKILL.mdfor skill creation guide
Built with ❤️ for AI-augmented productivity