39 installable skills for AI agents. Memory, security, governance, personas, infrastructure. By Melisia Archimedes.
Each skill auto-activates based on conversational context.
# Clone and copy skills to your project
git clone https://github.com/mizukaizen/hive-doctrine-skills.git
cp -r hive-doctrine-skills/skills/* .claude/skills/claude mcp add --transport http hive-doctrine https://hive-doctrine-mcp.vercel.app/mcp| Skill | Triggers When... |
|---|---|
a2a-protocol-explained |
Building agent-to-agent communication systems |
agent-compliance-101 |
Navigating compliance requirements for AI agents |
agent-cost-calculator |
Estimating or optimising AI agent operating costs |
agent-debugging-flowchart |
Debugging agent failures systematically |
agent-memory-architecture |
Designing memory systems for persistent agents |
agent-memory-decision-tree |
Choosing between memory system approaches |
agent-onboarding-checklist |
Deploying a new agent into production |
agent-rule-conflict-pattern |
Diagnosing agents that stop responding mid-thought |
agent-security-checklist |
Hardening agents before deployment |
agent-self-diagnostic-prompt |
Building self-diagnostic capabilities into agents |
agent-wallet-setup |
Setting up crypto wallets for agent payments |
ai-alignment-architecture |
Designing alignment into agent architecture |
app-niches-2026 |
Exploring profitable app niches for 2026 |
autonomous-agent-security |
Securing autonomous agent deployments |
config-gate-orphan-pattern |
Finding safety controls that exist on paper but never fire |
context-window-optimisation |
Optimising context window usage |
distributed-ai-safety |
Designing distributed AI safety architectures |
docker-dir-path-resolution |
Debugging Docker container path issues |
evaluate-your-agent |
Evaluating whether an agent is actually working |
five-multi-agent-architectures |
Choosing between multi-agent architecture patterns |
living-presence-protocol |
Making agents feel present, not scripted |
mcp-server-from-scratch |
Building an MCP server from zero |
mcp-tools-explained |
Understanding how agents discover and use MCP tools |
meta-prompt |
Writing prompts that generate better prompts |
model-selection-guide |
Choosing the right LLM for a specific task |
multi-agent-coordination |
Coordinating agents without a central controller |
on-device-vs-cloud-agents |
Deciding between on-device and cloud agent deployment |
prompt-injection-defence |
Defending against prompt injection attacks |
rag-vs-finetuning-vs-prompting |
Choosing between RAG, fine-tuning, and prompting |
redeemer-enum-mismatch |
Debugging pipelines that run perfectly but process nothing |
research-mega-prompt |
Turning any LLM into a structured research analyst |
silent-redemption-failure |
Diagnosing when systems report success but money never arrives |
soul-md-standard |
Understanding the SOUL.md identity specification |
soul-md-template |
Creating constitutional identity documents for agents |
stigmergic-vs-centralised-routing |
Choosing between coordination patterns |
system-prompt-patterns |
Applying proven system prompt patterns |
telegram-bot-block-handoff |
Implementing multi-agent chat handoff |
the-melissae |
Understanding the mythological foundation of AI safety |
what-is-stigmergy |
Learning about stigmergic coordination for agents |
/hive:browse — Browse the full product catalogue
/hive:align — Load the 7 alignment principles
/hive:search — Search the knowledge base
These skills are based on The Hive Doctrine — a thesis on polytheistic AI safety. Read the full thesis at hivedoctrine.com.
"The bee does not remember every flower. She remembers where the field is rich."
Skills: MIT License Content © 2026 Melisia Archimedes.