Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →aiskillstore/building-multiagent-systems
This skill should be used when designing or implementing systems with multiple AI agents that coordinate to accomplish tasks.
Passed 19 of 22 test cases with the skill, and 16 of 22 without it.
and 3 more in the full report
3 fixed16 passed either way3 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 3 test cases the skill fixed. Same model, same prompt, with and without the skill.
I am designing a multi-agent backend in Python to analyze complex financial documents and SEC filings with high reliability. Could you guide me through setting up this architecture? Please begin by asking any crucial scoping questions you need to understand my technical constraints and workflow, then provide a recommended blueprint detailing how to structure the system layers, choose the right agent orchestration strategy for near-zero error tolerance, and handle lifecycle safety like child process cleanup.
| Checked | Without skill | With skill |
|---|---|---|
| Asks the six mandatory discovery questions regarding starting point, primary use case, scale expectations, state requirements, tool coordination, and existing constraints. | Missed | Passed |
| Recommends the MAKER pattern for near-zero error tolerance requirements. | Missed | Passed |
| Structures the architecture using the four-layer stack consisting of Reasoning, Orchestration, Tool Bus, and Deterministic Adapters. | Missed | Passed |
The instructions your model reads when the skill loads.
Comprehensive architecture patterns for multi-agent systems where AI agents coordinate to accomplish complex tasks using tools. Language-agnostic and applicable across TypeScript, Python, Go, Rust, and other environments.
Before architecting any system, ask these six mandatory questions:
Every agent follows the four-layer architecture for testability, safety, and modularity:
| Layer | Name | Responsibility |
|---|---|---|
| 1 | Reasoning (LLM) | Plans, critiques, decides which tools to call |
| 2 | Orchestration | Validates, routes, enforces policy, spawns sub-agents |
| 3 | Tool Bus | Schema validation, tool execution coordination |
| 4 | Deterministic Adapters | File I/O, APIs, shell commands, database access |
Critical Rule: Everything below Layer 1 must be deterministic. No LLM calls in tools.
See references/four-layer-architecture.md for detailed implementation with code examples.
| Pattern | Purpose |
|---|---|
| Event-Sourcing | All state changes as events for audit trails and replay |
| Hierarchical IDs | Encode delegation hierarchy (e.g., session.1.2) for cost aggregation |
test cases passed, vs 16 / 22 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.