Build and demo Java AI agent systems with langchain4j-agentic: workflow patterns, supervisor, custom Planner strategies (incl. the flagship typed-verdict / CriticResult-style critic pattern), plus MCP tools, A2A remote agents, build setup, and conference-demo storylines. Pinned to 1.15.0 / 1.15.0-beta25.
84
89%
Does it follow best practices?
Impact
100%
4.76xAverage score across 2 eval scenarios
Passed
No known issues
A Tessl tile giving AI coding agents authoritative context for building and demoing Java AI agent systems with the langchain4j-agentic module. Combines official-docs API reference (langchain4j/docs/docs/tutorials/agents.md) with demo-building guidance, integrations, and runnable templates.
The centerpiece is the typed-verdict Custom Strategy (illustrated as CriticResult — a user-defined inter-agent object, not a framework type) — see docs/custom-strategy-critic-result.md.
| Type | Path | Loading |
|---|---|---|
| Skill | skills/langchain4j-agentic/SKILL.md | On invocation |
| Rules | rules/agentic-best-practices.md | Always loaded (steering) |
| Docs | docs/*.md | On-demand via MCP |
| Assets | skills/langchain4j-agentic/assets/ | build.gradle.kts, App.java |
| Scripts | skills/langchain4j-agentic/scripts/ | check_versions.sh |
docs/index.md — overview, current versions, navigationdocs/custom-strategy-critic-result.md — ⭐ flagship typed-verdict Custom Strategy, CriticResult-style (the "money slide")docs/agents-and-scope.md — @Agent, the builder, AgenticScopedocs/workflow-patterns.md — sequential, loop, parallel, parallel-mapper, conditional, human-in-the-loop, planner+executor+criticdocs/agent-configuration.md — optional, async, streaming, dynamic model, error handling, observabilitydocs/declarative-api.md — annotation API, supplier annotations, strongly typed keysdocs/pure-agentic.md — supervisor, custom Planner patterns (goal-oriented, P2P, voting)docs/tools-memory-and-build.md — tools, memory, models, -parameters, BOM split, dependenciesdocs/mcp-and-a2a.md — MCP tool servers and A2A remote agentsdocs/demo-storylines.md — four-act conference/workshop demodocs/gotchas.md — top-10 live-demo risks + API-correctness notesConfirmed current 2026-05-25 via maven-metadata.xml (refresh with scripts/check_versions.sh):
langchain4j / -anthropic / -bom: 1.15.0 (GA)langchain4j-agentic / -agentic-a2a / -agentic-patterns / -mcp: 1.15.0-beta25 (pin explicitly — not in BOM)tessl tile lint . # validate structure & content
tessl install file:. # install from filesystem into the current repo
tessl tile pack --output ./dist .Set a real workspace in tile.json (name: "<workspace>/langchain4j-agentic"), then tessl tile publish.
langchain4j-agentic is experimental; the BOM does not manage the beta artifacts — pin them independently.langchain4j-agentic-patterns.WIP/langchain4j-agentic-research/ folder holds the original demo research this tile was reconciled from; it is not part of the published tile.