UML Forge

Overview Schema Source Related Servers Discussions Admin

umlforge_ai_agent

Read-only

Design AI agent pipelines and multi-agent orchestration systems. Model agents, tool calls, memory, handoffs, and failure behavior as UML diagrams.

Instructions

Design an AI agent pipeline or multi-agent orchestration system.

USE THIS WHEN:

You are building a system where LLMs call tools or hand off to other agents
You want to visualise a multi-agent workflow (planner → researcher → writer)
You need to document tool access, memory strategy, and failure behaviour

NOT FOR:

General system architecture → use umlforge_stakeholder_arch
Standard synchronous API flows → use umlforge_api_sequence
Event-driven pipelines without LLM agents → use umlforge_event_driven

Produces:

Agent pipeline sequence: agents as participants with model names, tool calls as self-calls, human-in-the-loop gates, retry/fallback logic
Agent component map: agents, tool deps, memory, external integrations
Agent responsibility matrix: model, role, tools, inputs, outputs, failure behaviour
Risk & observability note: hallucination hotspots, validation gates, logging points
Tool overload flags (⚠️) for agents with more than 5 tools
(report_mode=True) Agent Pipeline Analysis Notes: pipeline risks, coverage gaps, reliability quick wins

Args: pipeline_purpose: What the agent system does (e.g. "Research pipeline that queries the web and drafts a report"). agents: Agents and their roles (e.g. "Planner [claude-opus-4], Researcher [claude-sonnet-4]"). tools_available: Tools agents can call (e.g. "web_search, execute_code") (optional). orchestration_approach: Coordination strategy — sequential, DAG, hierarchical, parallel fan-out (optional). memory_strategy: Memory approach — shared context, vector memory, Redis, none (optional). report_mode: True → also produce Agent Pipeline Analysis Notes. Pro/Team/Enterprise only. Omit to use your saved account preference (dashboard toggle); False forces diagrams only.

Input Schema

Name	Required	Description	Default
`agents`	Yes
`report_mode`	No
`memory_strategy`	No
`tools_available`	No
`pipeline_purpose`	Yes
`orchestration_approach`	No

Output Schema

Name	Required	Description	Default
`result`	Yes

Schema Changelog

Changes observed during successful MCP inspections.

Changed3 schema fields changedv0.3.0Oct 16, 2026
- addedInput schema / properties / report_mode / anyOf
```
Added value: +[
+  {
+    "type": "boolean"
+  },
+  {
+    "type": "null"
+  }
+]
```
- changedInput schema / properties / report_mode / default
```
Previous value: -falseNew value: +null
```
- removedInput schema / properties / report_mode / type
```
Removed value: -"boolean"
```
First observedv0.1.9Jul 21, 2026

TDQS

A4.7/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnlyHint=true, destructiveHint=false and openWorldHint=true, so the safety profile is covered. The description adds real context beyond that: the exact artifacts produced (sequence, component map, responsibility matrix, risk note), the ⚠️ tool-overload heuristic, and the plan-gating that report_mode requires Pro/Team/Enterprise. It does not explain persistence or whether generated diagrams are saved, which openWorldHint implies.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded and header-segmented so an agent can skim purpose, routing, outputs and args in order. The Produces list and Args block are lengthy, but with 0% schema coverage the per-parameter prose is doing necessary work rather than padding.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a six-parameter generation tool with an output schema, the description covers the required inputs, all optional inputs, the plan-tier constraint on report_mode, and the downstream artifacts. Nothing needed to invoke it correctly is missing, and return values are correctly left to the output schema.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0%, so the description carries the full burden and does it: each of the six parameters gets an explanation plus a concrete example, including which are optional and how report_mode's omission falls back to a saved account preference while False forces diagrams-only.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The opening line states a specific verb and resource ('Design an AI agent pipeline or multi-agent orchestration system') and the NOT FOR block names the exact sibling to use instead for each adjacent scenario, so an agent can route without opening any schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicit USE THIS WHEN / NOT FOR structure gives three concrete trigger scenarios and three named alternatives (umlforge_stakeholder_arch, umlforge_api_sequence, umlforge_event_driven) with the condition that selects each. Nothing is left to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server