Super ai ml agents

Super AI/ML Agents

Overview

Build reliable agent systems with clear tool boundaries, memory, and orchestration.

User Intent Examples

• "Need help with Agent Memory Systems for my product/site."
• "Create a plan for Agent Architecture."
• "Audit or improve Multi-Agent Patterns."

Workflow

1. Define the agent role, scope, and allowable actions.
2. Design tool interfaces and safety constraints.
3. Choose memory strategy: short-term, long-term, and retrieval.
4. Select orchestration pattern (single agent, multi-agent, supervisor).
5. Implement guardrails, fallbacks, and human-in-the-loop controls.
6. Validate behaviour with scenarios and regression tests.

Minimal Intake Questions

• Primary goal or outcome
• Scope (pages, systems, teams, or timeframe)
• Constraints (tools, budget, timeline)

Output Format

• Agent scope and behaviour contract
• Tooling and permissions plan
• Memory architecture
• Orchestration pattern and flow
• Validation plan with scenarios

Routing Map (Modules)

• Agent Memory Systems -> references/modules/agent-memory-systems.md
• Agent Architecture -> references/modules/ai-agents-architect.md
• Multi-Agent Patterns -> references/modules/multi-agent-patterns.md

Bundled References

• references/modules/
• scripts/
• assets/
• agents/

Compatibility Notes

• If any module references slash commands or tool-specific paths, translate them into plain-language steps.
• Keep outputs platform-agnostic unless the user specifies a specific tool, stack, or agent.

Guardrails

• Avoid ambiguous tool routing; keep responsibilities explicit.
• Keep memory bounded and purge stale data.
• Use human approval for high-risk actions.