kai-agents-skill
by rafaelcallejav1.1.0
Dynamic agent composition and orchestration system - create custom agents with unique personalities, voices, and trait combinations on-the-fly
1 skill GitHub
Documentation
# kai-agents-skill > Dynamic agent composition system - create specialized agents with unique personalities and voices, composed from traits on-the-fly Create custom agents by combining expertise, personality, and approach traits. 800 unique agent combinations possible. ## Features - **28 Composable Traits** - 10 expertise, 10 personality, 8 approach - **Hybrid Agent Model** - Named agents (persistent) + Dynamic agents (task-specific) - **Voice Mapping** - Each trait combination maps to a distinct voice - **AgentFactory CLI** - Compose agents programmatically with trait inference **Core Philosophy:** Cognitive diversity through composable agent personalities. Every analysis benefits from multiple perspectives. ## Installation ### Option 1: Marketplace (Recommended) ```bash # From Claude Code, run: /plugin marketplace add rafaelcalleja/kai-agents-skill /plugin install kai-agents-skill@rafaelcalleja-kai-agents-skill ``` ### Option 2: Development/Testing ```bash # Clone and use directly git clone https://github.com/rafaelcalleja/kai-agents-skill.git claude --plugin-dir ./kai-agents-skill ``` ### Option 3: CLI Tools Only ```bash cd kai-agents-skill/skills/Agents/Tools bun install ``` ## Verify Installation Once installed, test with: - Ask: *"create custom agents to research this topic"* → activates the agents skill - Ask: *"what agent traits are available?"* → lists all 28 composable traits - Ask: *"spin up custom security agents"* → creates specialized security agents ## Components ### Skill: Agents Provides dynamic agent composition. Triggers on: - "create custom agents", "spin up agents" - "specialized agents", "agent personalities" - "list available traits", "agent voices" ### Trait Categories | Category | Count | Options | |----------|-------|---------| | **Expertise** | 10 | security, legal, finance, medical, technical, research, creative, business, data, communications | | **Personality** | 10 | skeptical, enthusiastic, cautious, bold, analytical, creative, empathetic, contrarian, pragmatic, meticulous | | **Approach** | 8 | thorough, rapid, systematic, exploratory, comparative, synthesizing, adversarial, consultative | **Total combinations:** 10 x 10 x 8 = **800 unique agents** ## Usage ### Natural Language Just ask naturally: - "Create a custom security agent" - "I need a skeptical legal expert" - "Spin up 3 research agents with different approaches" ### AgentFactory CLI ```bash cd skills/Agents/Tools # List all available traits bun run AgentFactory.ts --list # Compose agent by traits bun run AgentFactory.ts --traits "security,skeptical,thorough" # Infer traits from task bun run AgentFactory.ts --task "Review this API for vulnerabilities" ``` ## Directory Structure ``` kai-agents-skill/ ├── .claude-plugin/ │ ├── marketplace.json │ └── plugin.json ├── skills/ │ └── Agents/ │ ├── SKILL.md │ ├── AgentPersonalities.md │ ├── Data/ │ │ └── Traits.yaml │ ├── Templates/ │ │ └── DynamicAgent.hbs │ ├── Tools/ │ │ └── AgentFactory.ts │ └── Workflows/ │ ├── CreateCustomAgent.md │ └── ListTraits.md └── README.md ``` ## Credits - **Origin:** Based on [kai-agents-skill pack](https://github.com/danielmiessler/Personal_AI_Infrastructure) by Daniel Miessler - **License:** MIT ### Acknowledgments - **Daniel Miessler** - Original PAI architecture and agent concepts - **IndyDevDan** - Trait composition patterns and voice mapping - **Anthropic** - Claude 4.x agent capabilities ## Changelog ### v1.1.0 (2026-01-01) Synced with standard plugin structure: - **BREAKING**: Moved `plugin.json` to `.claude-plugin/` directory - Updated paths to use `CLAUDE_PLUGIN_ROOT` for portability - Custom agents now use `subagent_type: "general-purpose"` instead of "Intern" - Removed `.claude/settings.local.json` ### v1.0.0 - Initial release - 28 composable traits (10 expertise, 10 personality, 8 approach) - AgentFactory CLI tool with trait inference - Hybrid agent model (named + dynamic)