Template library

Start from code that runs.

Only templates marked ✓ Tested have passed an end-to-end run. “Coming soon” cards are the roadmap, not finished work.

OpenAI Agents SDK · v0.24.0 · Python✓ Tested

Approval + resume & tool error formatting

Pause tool calls for human approval, checkpoint the run to a string, approve/reject, and resume. Plus the new run-wide tool_error_formatter that scrubs secrets from crashed tools.

#human-in-the-loop#durability#error handlingTested on openai-agents 0.24.0Free

LangGraph · 1.2.15 · Python✓ Tested

Human-in-the-loop approval graph

A propose → approval → execute StateGraph that pauses with interrupt(), saves the pause in a checkpointer under a thread_id, and resumes with Command(resume=...) down the approve or reject path.

#human-in-the-loop#checkpointingTested on langgraph 1.2.15Free

LangGraph · 1.2.15 · Python✓ Tested

Idempotent approval: no double charges on resume

When a run resumes after interrupt(), LangGraph re-runs the paused node from the top. This template shows the double-charge bug and the fix: approval-only node, separate execute node, and an idempotency key derived from thread_id.

#human-in-the-loop#idempotency#checkpointingTested on langgraph 1.2.15Pro

OpenAI Agents SDK · v0.24.0 · Python✓ Tested

Idempotent approved side effects

The Agents SDK runs an approved tool once per resume, but side effects still repeat on a duplicate resume of the same RunState or a whole-run retry after a crash. Shows both failures (2 charges) and the fix (1 charge).

#human-in-the-loop#idempotency#durabilityTested on openai-agents 0.24.0Pro

LangGraph · 1.2.15 · Python✓ Tested

Multiple interrupts in one node

One node calls interrupt() three times and is answered one resume at a time (matched by position), plus two parallel nodes that interrupt in the same step and are resumed together with a dict keyed by Interrupt.id.

#human-in-the-loop#interruptsTested on langgraph 1.2.15Free

LangGraph · 1.2.15 · Python✓ Tested

Minimal web UI to approve paused runs

A stdlib-only web page (plain HTML forms, no JavaScript) that lists paused LangGraph runs and approves or rejects them with Command(resume=...), with 409/400/404 handling.

#human-in-the-loop#approval UITested on langgraph 1.2.15Free

Vercel AI SDK · 7.0.142 · TypeScript✓ Tested

Tool approval with generateText

generateText with the v7 toolApproval option: refunds over $100 pause as tool-approval-request, a human approves or denies, and the run resumes. Covers $50 auto-approved, $500 approved and $500 denied.

#human-in-the-loop#tool approvalTested on ai 7.0.142Free

CrewAI · PythonSoon

Research crew with sourced output

Planned: a role-based crew (researcher, writer, reviewer) that returns a cited brief. Not built or tested yet.

#multi-agentPro

Google ADK · 2.12.0 · Python✓ Tested

Tool confirmation step

An ADK agent whose transfer_money tool asks a human to confirm only large amounts (require_confirmation as a callable), then runs or rejects the original call in the same session.

#human-in-the-loop#tool confirmationTested on google-adk 2.12.0Free

Pydantic AI · PythonSoon

Durable, typed task runner

Planned: typed agent outputs with a task that survives restarts. Not built or tested yet.

#durability#typed outputsPro

MCP · TypeScriptSoon

MCP server starter

Planned: a minimal Model Context Protocol server exposing two tools, with a local test client. Not built or tested yet.

#tools#MCPFree

Cloudflare Agents · TypeScriptSoon

Scheduled agent on the edge

Planned: a stateful agent that schedules its own follow-up tasks. Not built or tested yet.

#schedulingPro