Claude

Use Agenisea's Agent Designer
with Claude Code

Design multi-agent infrastructure with topology, orchestration, and resilience directly from your terminal using Claude Code.

Agenisea is not a code generator. It's an infrastructure design aid for thinking clearly about agent systems before implementation.

Download instructions for offline reference.

Installation

Option 1: Install globally (available in all projects)

curl -o ~/.claude/commands/agenisea.md https://agenisea.ai/agenisea.md

Option 2: Install per-project

mkdir -p .claude/commands && curl -o .claude/commands/agenisea.md https://agenisea.ai/agenisea.md

Option 3: Manual download

Download agenisea.md and save it to ~/.claude/commands/ or .claude/commands/

Usage

Run the slash command:

/agenisea

Then describe your multi-agent workflow:

What: A code review pipeline that analyzes PRs for bugs, security issues, and style

Agents: 4 specialized reviewers (security, logic, style, summary)

Pattern: Parallel analysis, then sequential synthesis

Constraints: 50k token budget, 30s timeout per agent

Resilience: Fall back to single general reviewer if specialists fail

What you get

Agenisea creates a complete multi-agent architecture blueprint:

  • 1.
    Agent Topology — Role, inputs, outputs, and dependencies for each agent
  • 2.
    Pipeline Orchestration — Sequential, parallel, or conditional execution flow
  • 3.
    Resilience Patterns — Circuit breakers, token budgets, timeouts, fallback chains
  • 4.
    Agent Contracts — System prompts with handoff protocols and guardrails
  • 5.
    Implementation Scaffold — Directory structure and module breakdown
  • 6.
    Recommended Tech Stack — Research-backed framework and library recommendations

Execution patterns

  • Sequential — Agents execute in order, each waits for previous. Use when output of one is input to next.
  • Parallel — Agents execute simultaneously, results merged. Use for independent subtasks.
  • Conditional — Route based on input or intermediate results. Use for different paths for different scenarios.
  • Hybrid — Combination of above. Use for complex workflows with branches and joins.

Agent role templates

Common agent archetypes to use in your blueprints:

  • Researcher — Gather information from sources
  • Analyzer — Extract patterns, insights, structure
  • Validator — Check correctness, completeness, safety
  • Synthesizer — Combine multiple inputs into coherent output
  • Router — Decide which path/agent to invoke
  • Critic — Review and suggest improvements
  • Executor — Take actions (API calls, writes, etc.)
  • Monitor — Track progress, detect issues, report status