Agentic LangGraph — From First Principles to a Working Multi-Agent System

A new learning repository, Agentic-Langgraph-custom, extends Krish Naik's Agentic LangGraph Crash Course with a custom multi-agent module, demonstrating a two-agent research-to-report pipeline using openai/gpt-oss-20b served by Groq, Tavily for search, and LangSmith for observability. The repository includes detailed troubleshooting guidance, such as handling rate-limit pauses and correctly configuring the LangSmith endpoint for AWS data regions, based on real runs.

A learning repository that goes all the way from “what is a node?” to a two-agent research → report pipeline, fully traced in LangSmith and debuggable in LangGraph Studio. Every code snippet here is taken from the modules in this repo, and every number, error message and screenshot comes from a real run on this machine — not from documentation. 8. Troubleshooting — every error we actually hit https://github.com/faizulkhan56/Agentic-Langgraph-custom 8-troubleshooting--every-error-we-actually-hit A fork of Krish Naik’s Agentic LangGraph Crash Course , extended with a custom multi-agent module and a lot of hard-won operational detail. Model used throughout: openai/gpt-oss-20b served by Groq. Search: Tavily. Observability: LangSmith AWS data region . git clone <your-repo-url cd Agentic-Langgraph-custom 1. install uv is the source of truth; uv.lock is committed uv sync 2. credentialscp .env.example .env then edit .env — see section 3 3. run the multi-agent pipelinecd 5-Multitoolspython agent.py Expected: 2–3 minutes of wall clock, a printed flow, and a markdown report in 5-Multitools/reports/. If it looks stuck need to read section 8.4 https://github.com/faizulkhan56/Agentic-Langgraph-custom 84-why-does-a-run-take-3-minutes — it is almost certainly rate-limit pauses, not a hang. To explore visually instead: cd 5-Multitools && langgraph dev opens LangGraph Studio This is the part that silently breaks everything, so it comes before the theory. Settings → API Keys. A key is shown once at creation — if we lose it, create a new one. Note the Personal / Service toggle: personal tokens are scoped to only us, service keys to a workspace. --- LLM provider ---GROQ API KEY="gsk ..." optional: override the model used by the agents GROQ MODEL="openai/gpt-oss-20b" --- web search ---TAVILY API KEY="tvly-..." --- LangSmith tracing ---LANGCHAIN API KEY="lsv2 pt ..." Data region — see 3.4. THIS ONE MATTERS.LANGSMITH ENDPOINT="https://aws.api.smith.langchain.com" optional LANGSMITH PROJECT="TestProject" .env is gitignored. .env.example is committed as the template. Never commit .env — verify with git check-ignore -v .env before your first push. Three consumers, one file: { "dependencies": "." , "graphs": { "multi agent": "./agent.py:multi agent" }, "env": "../.env"} Then the code promotes and validates them: php load dotenv def require name: str - str: value = os.getenv name if not value: raise RuntimeError f"{name} is missing. Add it to the .env at the repo root." return valueos.environ "GROQ API KEY" = require "GROQ API KEY" require "TAVILY API KEY" langsmith key = os.getenv "LANGSMITH API KEY" or os.getenv "LANGCHAIN API KEY" if langsmith key: os.environ "LANGSMITH API KEY" = langsmith key os.environ "LANGSMITH TRACING" = "true" os.environ "LANGSMITH PROJECT" = "multiagent-summarizer" os.environ.setdefault "LANGSMITH ENDPOINT", "https://aws.api.smith.langchain.com" Why require instead of the common os.environ "X" = os.getenv "X" ? Because if the variable is missing, that idiom raises TypeError: str expected, not NoneType from deep inside os.environ, which tells you nothing. require names the variable and the file. Why LANGSMITH API KEY or LANGCHAIN API KEY? The SDK renamed its variables; both are still honoured, LANGSMITH wins. This repo stores the key under the older name and promotes it, so either works. Why setdefault for the endpoint but = for the project? setdefault lets .env override the region portable for other users , while = guarantees this module's traces never leak into another module's project. LangSmith runs three separate deployments. A key issued in one is rejected by the others with an error that looks exactly like an invalid key. LangChain Support confirming the account is on the AWS region and must set LANGSMITH ENDPOINT. Symptom if you get this wrong: Failed to multipart ingest runs: langsmith.utils.LangSmithError:Failed to POST https://api.smith.langchain.com/runs/multipart in LangSmith API.HTTPError '403 Client Error: Forbidden ...', '{"error":"Forbidden"}' Critical gotcha: LangSmith returns 403 for unknown keys too — not 401. So a 403 tells you "auth failed" and nothing more. It does not mean "the key is recognised but restricted". Do not conclude your key was revoked. Find your region in one command: k=$ grep '^LANGCHAIN API KEY' .env | cut -d= -f2- | tr -d '"'"'"' \r' for h in api eu.api aws.api; do printf "%-8s - " "$h" curl -s -o /dev/null -w "%{http code}\n" -H "x-api-key: $k" \ "https://$h.smith.langchain.com/api/v1/sessions?limit=1"done 200 on exactly one host is your endpoint. The tr -d also strips the trailing \r that Windows CRLF line endings leave in the value — leave it in and the header itself is malformed. After editing .env, restart the kernel. Two reasons: So re-running the load dotenv cell changes nothing, and you keep hitting the same 403 with a correct .env sitting on disk. LangChain gives us a chain: a fixed pipeline, A → B → C. LangGraph gives us a state machine: nodes that read and write they all shared a state, and edges — including conditional ones — that decide where to go next. That is what makes loops possible, and a loop is what turns a chatbot into an agent. Three concepts and nothing else: ┌─────────── STATE ───────────┐ a dict every node reads and writes │ │ ── NODE ──► NODE ──► NODE ──► │ functions that return partial updates │ ▲ │ │ └── EDGE ─┘ EDGE ┘ │ wires, some conditional State is a TypedDict. Each key is a channel. A node returns a partial update; LangGraph merges it in. python from typing extensions import Annotated, NotRequired, TypedDictfrom langchain core.messages import BaseMessagefrom langgraph.graph.message import add messagesclass ReportState TypedDict : messages: Annotated list BaseMessage , add messages reducer: append research notes: NotRequired str optional, overwrite report: NotRequired str Reducers — the key idea Without a reducer, a returned value replaces the channel. With a reducer, it is merged. add messages is the reducer you will use most. Verified behaviour: That last row is how weedit history rather than grow it. NotRequired marks a key optional, which keeps Studio's input form clean: messages shows as required, the rest are blank until a node fills them. A node is any callable state - partial update. python def research agent state: ReportState : reply = research llm.invoke SystemMessage content=RESEARCH PROMPT , state "messages" update = {"messages": reply } merged by add messages if not getattr reply, "tool calls", None : update "research notes" = reply.content plain overwrite return update Three things worth copying from this: python from langgraph.graph import START, END, StateGraph builder = StateGraph ReportState builder.add node "research agent", research agent builder.add node "research tools", ToolNode research tools builder.add edge START, "research agent" entry pointbuilder.add edge "research tools", "research agent" unconditional A router function inspects state and returns the name of the next node. builder.add conditional edges "research agent", tools condition, the router {"tools": "research tools", END: "report agent"}, verdict - destination tools condition is LangGraph's prebuilt router. Verified signature: php def tools condition state, messages key="messages" - Literal "tools", " end " : ... It returns "tools" if the last AIMessage has tool calls, else " end ". The third argument is a path map, and it is the trick that makes multi-agent work. tools condition only ever says "tools" or END, but you get to decide what those words mean for each node. Here END from research agent is remapped to report agent — so "agent 1 has stopped calling tools" becomes "hand off to agent 2" instead of "finish" . One prebuilt router, two different meanings. graph = builder.compile validates the graph, returns a Runnablestate = graph.invoke {"messages": "Hi"} run to completion compile is where LangGraph checks that every edge points somewhere real. If you edit a node function afterwards, see the staleness warning https://github.com/faizulkhan56/Agentic-Langgraph-custom 83-i-edited-a-node-and-nothing-changed . Visualise it: python from IPython.display import Image, displaydisplay Image graph.get graph .draw mermaid png A tool is a Python function the model may call. @tool turns the signature and docstring into a JSON schema, so the docstring is a prompt — write it for the model. python from langchain core.tools import tool php @tooldef multiply a: int, b: int - int: """Multiply a and b. Args: a int : first int b int : second int Returns: int: output int """ return a b Bind them to the model and put them in a ToolNode: llm with tools = llm.bind tools multiply, addition, get weather, search builder.add node "tools", ToolNode multiply, addition, get weather, search bind tools tells the model what it may call. ToolNode is what actually executes the call and appends a ToolMessage. You need both. Binding TavilySearch directly returns full page content per hit — measured at 8,130 tokens for one call, which alone exceeded the token budget. Wrapping it fixes that and costs nothing: php @tooldef web search query: str - str: """Search the web for current information on a topic. Returns a compact list of results: title, URL and a short snippet each. """ raw = TavilySearch max results=4 .invoke {"query": query} blocks = for hit in raw.get "results", : snippet = " ".join hit.get "content" or "" .split :280 blocks.append f"- {hit.get 'title' }\n url: {hit.get 'url' }\n {snippet}" return "\n".join blocks or f"No results for: {query}" In the trace this shows as tavily search nested inside web search — proof the wrapper is in the path. Sometimes a tool needs data the model shouldn’t have to retype. Make the model pass the report text to three tools and it crosses the wire four times per run. python from langgraph.prebuilt import InjectedState php @tooldef report stats state: Annotated dict, InjectedState - dict: """Measure the report already drafted by make markdown report. Takes no arguments.""" text = latest draft state "messages" words = len text.split return {"words": words, "characters": len text , "bullets": text.count "\n- " , "reading time minutes": max 1, round words / 200 } InjectedState is filled in by LangGraph at execution time and is absent from the schema the model sees. In the trace you will see report stats called with no arguments. This one change removed roughly two-thirds of the token cost of agent 2. Reason + Act: the model calls a tool, reads the result, and decides again. In LangGraph that is a single edge pointing backwards. builder.add conditional edges "agent", tools condition builder.add edge "tools", "agent" ← the loop Contrast: WiringBehaviourbuilder.add edge "tools", END one tool call, then stop builder.add edge "tools", "agent" model reads the result and may call again Why it matters, from a real run of “What is 12 multiplied by 4, then add 8?” : php AI - multiply a=12, b=4 TOOL - 48AI - addition a=48, b=8 ← 48 came from the previous tool resultTOOL - 56AI - "The result is 56." multiply → 48, then addition a=48, b=8 → 56. Without the back-edge the run stops after multiply and the second step never happens. Models like to shortcut trivial work. A system prompt fixes it: SYSTEM PROMPT = "For every distinct action requested weather, search, math, etc. , you must " "call the corresponding tool - never compute or answer any part yourself, " "even if it seems trivial." With this in place, a three-part request routed all three parts through tools: “Give me the weather of dhaka and then give me recent sports news finally add 5 with 4” → get weather, tavily search, and addition — the model called a tool even for 5 + 4. invoke returns the final state. stream yields after each node, which is how you watch an agent work. for event in graph.stream {"messages": "Write a brief report on EV battery recycling"} : for node, value in event.items : last = value "messages" -1 calls = c "name" for c in getattr last, "tool calls", None or print f"{node:<16} tools={calls or '-'}" research agent tools= 'web search' research tools tools=-research agent tools= 'web search' research tools tools=-research agent tools=- ← handoffreport agent tools= 'make markdown report' By default a graph is stateless — every invoke starts empty: graph.invoke {"messages": "Hello my name is faizul"} graph.invoke {"messages": "What is my name"} no idea Add a checkpointer and a thread: python from langgraph.checkpoint.memory import MemorySaver memory graph = builder.compile checkpointer=MemorySaver config = {"configurable": {"thread id": "1"}} r1 = memory graph.invoke {"messages": "Write a short report on Bangladesh IT outsourcing"}, config=config r2 = memory graph.invoke {"messages": "Now add a section on the main risks"}, config=config r2's prompt never names the topic, and it still works — the checkpointer replayed thread "1". For persistence across restarts use SqliteSaver or PostgresSaver instead. Under langgraph dev, leave the checkpointer out — the platform supplies its own, which is why the factory takes it as an argument: python def make multi agent graph checkpointer=None : ... return builder.compile checkpointer=checkpointer multi agent = make multi agent graph Studio: no checkpointer Three things about memory that will bite you 1. Token cost compounds. Replaying the thread means every turn is bigger. Measured across one run: Eventually a single call exceeds your per-minute budget and you get 413 Request too large, which no retry can fix. Use a fresh thread id per topic: config = {"configurable": {"thread id": "2"}} new topic, clean history 2. Follow-ups inherit the original framing. Asking “Now add a section on the AI risks” produced a document still titled “Short Note on AI Definition” — agent 2 re-drafts the whole thread and titles it after turn one. 3. Each result is an independent snapshot. response, r1 and r2 are separate dicts. Printing response 'report' after a memory run shows the old document. Nothing updates a variable you assigned earlier. LangSmith is the difference between “it broke” and “it broke on the fourth LLM call inside agent 2, here are the exact arguments” . Everything in this section was used to debug the real failures in section 8 https://github.com/faizulkhan56/Agentic-Langgraph-custom 8-troubleshooting--every-error-we-actually-hit . Three environment variables, and tracing is automatic — no code changes, no wrappers: os.environ "LANGSMITH TRACING" = "true" the master switchos.environ "LANGSMITH API KEY" = "lsv2 pt ..."os.environ "LANGSMITH PROJECT" = "multiagent-summarizer" where traces landos.environ "LANGSMITH ENDPOINT" = "https://aws.api.smith.langchain.com" Every langchain / langgraph call is instrumented, so the graph, each node, each LLM call and each tool call are all reported. Set LANGSMITH TRACING=false to turn it off. Tracing failures are non-fatal — ingest runs on a background thread, so a 403 spams stderr but your graph still returns a correct answer. You lose the trace, not the result. os.environ "LANGSMITH PROJECT" = os.getenv "LANGSMITH PROJECT MULTIAGENT", "multiagent-summarizer" Plain = rather than setdefault so a stray global can't merge these traces into another module's project. This repo uses TestProject for 3-Debugging and multiagent-summarizer for 5-Multitools. Per-project rollups: trace count, error rate, P50/P99 latency, tokens, cost. The name of the root run tells you the entry point: Same graph, different launcher. Do not read anything else into it. New accounts land in an application scope that hides workspace projects: “No tracing projects in ‘My First App’ — 1 tracing project exists in your workspace, but it’s hidden because you’re viewing ‘My First App’.” Click All applications. Your project was there the whole time. tool agent 1.23s 1.2K tok $0.0002├─ tool calling llm 0.56s│ ├─ ChatGroq openai/gpt-oss-20b 0.55s 579 tok│ └─ tools condition 0.00s├─ tools 0.01s│ └─ multiply 0.00s└─ tool calling llm 0.54s └─ ChatGroq 0.53s 592 tok How to read it: The two-agent pipeline: research agent → tools condition → research tools → web search → tavily search, then the handoff to report agent. The Attributes → Metadata panel also shows the LANGSMITH ENDPOINT and LANGSMITH PROJECT actually in effect — useful when you are unsure which config the process picked up. The single most useful debugging technique in this repo. A malformed tool call was crashing runs; a retry was added; runs still crashed. Was the retry running? Count the LLM spans under the failing node. The real trace: 10:39:40 report agent error TOOL USE FAILED10:39:40 ChatGroq error TOOL USE FAILED ← ONE span One span proved the kernel was running pre-patch code — the fix was correct, the process was stale. Without this the obvious and wrong conclusion is “the retry doesn’t work”. Counter-example: token count is not a retry signal. A failing run showed 2,861 tokens, which looks like several attempts. It was three successful research calls; the failed call billed 0 tokens. Count spans, not tokens. The UI is for browsing; the API is for answering questions. These cost no LLM tokens — invaluable when you are rate-limited. k=$ grep '^LANGCHAIN API KEY' .env | cut -d= -f2- | tr -d '"'"'"' \r' BASE=https://aws.api.smith.langchain.com Does my project exist and when did it last run? curl -s -H "x-api-key: $k" "$BASE/api/v1/sessions?name=multiagent-summarizer" What failed, and why? sid=$ curl -s -H "x-api-key: $k" "$BASE/api/v1/sessions?name=multiagent-summarizer" \ | python -c "import sys,json;print json.load sys.stdin 0 'id' " curl -s -X POST "$BASE/api/v1/runs/query" -H "x-api-key: $k" \ -H "Content-Type: application/json" \ -d "{\"session\": \"$sid\" ,\"is root\":true,\"limit\":20, \"select\": \"name\",\"status\",\"error\",\"start time\",\"end time\",\"total tokens\" }" Real output, summarised: php 10:55:04 - 10:57:36 success tok=2742510:53:16 - 10:53:45 success tok=1109310:52:24 - 10:52:25 error tok=0 Drop "is root": true to get every span, which is how you count LLM calls per node 5.5 https://github.com/faizulkhan56/Agentic-Langgraph-custom 55-the-span-count-trick--proving-a-retry-is-live . Is my key valid, and for which region? See 3.4 https://github.com/faizulkhan56/Agentic-Langgraph-custom 34-the-data-region-trap--read-this-before-you-debug-a-403 . Monitoring → Cost & Tokens: totals and per-trace tokens over time. Measured on this repo: Latency profile for the single-agent tools: A visual debugger for our graph: run it, watch nodes light up, inspect state at each step. langgraph.json next to our agent.py: { "dependencies": "." , "graphs": { "multi agent": "./agent.py:multi agent" }, "env": "../.env"} The graphs value is module:variable — the variable must be a compiled graph at import time, which is why agent.py ends with multi agent = make multi agent graph . cd 5-Multitoolslanggraph dev or ../.venv/Scripts/langgraph.exe dev Healthy startup: Importing graph profiling graph id=multi agent path=./agent.pyApplication started up in 3.363sHTTP Request: POST https://aws.api.smith.langchain.com/v1/metadata/submit "204 No Content" That last line confirms Studio picked up your region from .env via langgraph.json Graph on the left, thread on the right. Type into messages and hit Submit. What to look for: Note that Studio talks to the same LangSmith project, so everything we do appears in your traces too — under the root-run name multi agent rather than LangGraph. langgraph dev runs in-memory. The bundled langgraph-verify-graphs tool targets the production server and needs Redis: KeyError: "Config 'REDIS URI' is missing, and has no default." That is expected, not a broken install. Verify a graph loads by importing it instead: python python -c "from agent import multi agent; print list multi agent.get graph .nodes " Everything above, assembled. The user asks about a topic. A search agent gathers information with Tavily. That information is handed to a summarizer/report agent, which has its own tools for summarizing, formatting and saving a polished report. ┌──────────────── shared messages channel ───────────────┐ │ │ START ──► research agent ⇄ research tools report agent ⇄ report tools ──► END │ web search ▲ make markdown report, │ │ report stats, └───── handoff: research notes ─────┘ save report builder.add edge START, "research agent" agent 1 loops; when it stops calling tools, hand offbuilder.add conditional edges "research agent", tools condition, {"tools": "research tools", END: "report agent"}, builder.add edge "research tools", "research agent" agent 2 loops; when it stops calling tools, we are donebuilder.add conditional edges "report agent", tools condition, {"tools": "report tools", END: END}, builder.add edge "report tools", "report agent" Two mechanisms, deliberately: Each agent stays in character because each node prepends its own system prompt to the same shared history. State keyContainsSame as the .md file?research notesagent 1's raw bullet notes + URLsnoreportthe make markdown report outputyes, identical text update "report" = latest draft state "messages" or reply.content report is deliberately not reply.content: agent 2's final message is a one-line sign-off, because the prompt forbids repeating the report a token-saving measure . So the document is recovered from the tool output instead. As a script: cd 5-Multitoolspython agent.py default topicpython agent.py Write a short report on AIOps in Bangladesh args are joined From a notebook: 5-Multitools/multiagent.ipynb, 41 cells, built up in the same order as section 4 https://github.com/faizulkhan56/Agentic-Langgraph-custom 4-langgraph-fundamentals . In Studio: langgraph dev, then type into messages TOPIC: Write a short report on electric vehicle battery recycling--- flow --- agent - calls web search tool web search: - ReCell In The News 2024 - ReCell Center agent - calls web search tool web search: - Future of EV Battery Recycling: Technologies and Challenge agent - text 3085 chars ← agent 1 done, handoff agent - calls make markdown report tool make markdown report: Electric Vehicle Battery Recycling: Current... agent - calls report stats tool report stats: {"words": 308, "characters": 2587, "bullets": 10, ...} agent - calls save report tool save report: Saved 2587 characters to reports/ev-battery-recycling-report.md agent - text 55 chars ← one-line sign-off Every design decision is visible here: two searches then a handoff, report stats called with no arguments InjectedState , save report given only a filename, and a 55-character sign-off instead of a repeated report. Produced document: Electric Vehicle Battery Recycling: Current Landscape and Future Outlook Compiled 2026-09-03 Executive summary... Key findings- Market size & growth — valued at USD 25.49 billion in 2024...- Current commercial recovery rates: ~70-80 % cobalt, ~60-70 % nickel... Sources1. ReCell Center 2024 — https://recellcenter.org/newsroom/recell-in-the-news-20242. U.S. DOE Fact Sheet 2024 — https://www.energy.gov/... Fail fast on config. require names the missing variable and the file. Degrade, don’t crash. Tavily is only wired in when its key exists, so the graph still loads for someone without a Tavily account: php def build research tools - list: tools = ... if os.getenv "TAVILY API KEY" : tools.insert 0, ... return tools Timeouts and readable failures on every HTTP tool — a raised exception inside a node kills the graph, a returned string lets the model recover: try: ... the HTTP callsexcept requests.RequestException as exc: return f"Weather lookup failed for {city}: {exc}" Constrain file writes. save report takes a filename from a language model, so it strips anything path-like and forces the extension: stem = re.sub r" ^A-Za-z0-9. - ", "-", Path filename .stem .strip "-." or "report"target = REPORTS DIR / f"{stem}.md" Retry what is retryable, and only that — see 8.1 https://github.com/faizulkhan56/Agentic-Langgraph-custom 81-400-tool use failed and 8.2 https://github.com/faizulkhan56/Agentic-Langgraph-custom 82-429-rate-limit . Console encoding on Windows. Reports contain U+2011 non-breaking hyphen , U+202F narrow no-break space and ≈. The console is cp1252, so without this the script dies with UnicodeEncodeError after spending all the tokens: sys.stdout.reconfigure encoding="utf-8", errors="replace" BadRequestError: Error code: 400 - Failed to parse tool call arguments as JSON'failed generation': '{"name": "web search", "arguments": {"query":"Bangladesh overview 2024","}""}' The model emitted invalid JSON for its own tool arguments. Observed corruption varies between runs — ,"" , ,"}"" — which is what proves it is random sampling damage rather than a schema bug. Confirmed intermittent: many runs succeed on the identical schema. Fix: re-sample. Do not switch models. RateLimitError: 429 - TPM: Limit 8000, Used 7841, Requested 4974.Please try again in 36.1125s Groq’s free tier allows 8,000 tokens per minute across all calls. Groq tells us how long to wait, so honour it. RETRY ATTEMPTS = 4MAX RATE LIMIT WAIT = 90.0 python def invoke with retry model, messages : """Retry the two failures this stack actually produces.""" last error = None for attempt in range 1, RETRY ATTEMPTS + 1 : try: return model.invoke messages except BadRequestError as exc: if "tool use failed" not in str exc : raise a real bug — surface it print f" malformed tool call: re-sampling {attempt}/{RETRY ATTEMPTS} " last error = exc except RateLimitError as exc: match = re.search r"try again in \d. + s", str exc wait = min float match.group 1 + 1.0, MAX RATE LIMIT WAIT if match else 20.0 print f" rate limited: waiting {wait:.0f}s, attempt {attempt}/{RETRY ATTEMPTS} " time.sleep wait last error = exc raise last error The if "tool use failed" not in str exc : raise line matters. Without it this swallows genuine 400s — a bad model name, a malformed request — and retries them pointlessly before failing with a confusing error. Verified behaviour: ScenarioResult2 malformed calls then successrecovers on attempt 3always malformed4 attempts, then raisesunrelated 400 bad model name re-raised immediately It deliberately does not handle 413 Request too large — waiting cannot shrink an oversized request. If you see 413, cut tokens or start a new thread. The most expensive hour in this repo. add node captures the function object: cell A invoke with retry ← helpercell B research agent / report agent ← closures calling itcell C builder.add node ... ← CAPTURES the function objectscell D builder.compile ... Re-running B rebinds the name but leaves the original function inside builder. graph and memory graph both inherit the stale node. Restart the kernel and Run All, or re-run A → B → C. Then verify for free: python import inspectfor name in "research agent", "report agent" : fn = graph.nodes name .node.steps 0 .func print f"{name:<15} retry wired: {'invoke with retry' in inspect.getsource fn }" The node is a RunnableSeq: steps 0 is your function, steps 1: are LangGraph's channel-write and routing steps. .node.func raises AttributeError. It is not a hang. Verified trace 10:55:04 → 10:57:36, success, 27,425 tokens: LLM callTokensGap before next step13,599–24,29631s34,77539s44,81733s54,85641s At ~4,800 tokens per call against 8,000 per minute you get roughly 1.6 calls per minute, and a run needs 6–8 calls. The arithmetic is the runtime. rate limited: waiting 37s, attempt 1/4 in the output is our liveness signal. No such lines and no progress for minutes is a real hang. To cut cost, at the top of agent.py: SEARCH MAX RESULTS = 3 from 4SEARCH SNIPPET CHARS = 180 from 280 plus RESEARCH PROMPT: "Run at most 2 searches" → "Run exactly 1 search" . Roughly 6–8k tokens per run, which fits one minute. Tradeoff: thinner research. 413 - Request too large ... TPM: Limit 8000, Requested 8130 A single request exceeded the budget. Binding TavilySearch directly caused this full page content per hit . Wrap noisy tools 4.5 https://github.com/faizulkhan56/Agentic-Langgraph-custom 45-tools--giving-the-model-hands and use InjectedState so the model never re-emits large text. Wrong data region. See 3.4 https://github.com/faizulkhan56/Agentic-Langgraph-custom 34-the-data-region-trap--read-this-before-you-debug-a-403 . Remember 403 also means “unknown key” — it is not evidence your key is valid. Windows cp1252 console meeting model output. Either sys.stdout.reconfigure encoding="utf-8", errors="replace" in the script, or: PYTHONIOENCODING=utf-8 python your script.py Notebooks are unaffected. PowerShell syntax in Git Bash. Backslash is an escape character there: ../.venv/Scripts/langgraph.exe dev Git Bash ..\.venv\Scripts\langgraph.exe dev PowerShell Agentic-Langgraph-custom/├── .env your keys — GITIGNORED, never commit├── .env.example template, committed├── .gitignore├── README.md this file├── pyproject.toml uv-managed deps├── uv.lock│├── 1-BasicChatbot/chatbot.ipynb├── 2-HumanAssistance/humanintheloop.ipynb├── 3-Debugging/│ ├── agent.py tool agent: search, weather, multiply, addition│ ├── debugging.ipynb│ └── langgraph.json├── 4-Multimodal/├── Agents/multiaiagent.ipynb│├── 5-Multitools/ ← the multi-agent project│ ├── agent.py multi agent graph + CLI runner│ ├── multiagent.ipynb 41-cell walkthrough│ ├── langgraph.json│ └── reports/ generated .md files gitignored │└── docs/ ├── SCREENSHOT-NOTES.md └── assets/langsmith-studio/ 26 annotated screenshots git check-ignore -v .env must report a matchgit ls-files | grep -i "\.env$" must return nothing .gitignore covers .env except .env.example , .venv/, pycache /, .langgraph api/ which holds real conversation state , .pckl, and 5-Multitools/reports/ .md. Notebooks store outputs. They are clean today, but a cell that prints os.environ would bake a live key into a committed .ipynb. To enforce: uv add --dev nbstripout && nbstripout --install The screenshots in docs/assets/ contain a LangSmith org UUID and a partial API key prefix. Neither is exploitable, but make publishing them a deliberate choice. Forked from Krish Naik’s Agentic LangGraph Crash Course and extended with the 5-Multitools multi-agent module, the LangSmith operational material, and the troubleshooting sections — all derived from real failures on this machine. Git link: faizulkhan56/Agentic-Langgraph-custom https://github.com/faizulkhan56/Agentic-Langgraph-custom Agentic LangGraph — From First Principles to a Working Multi-Agent System https://pub.towardsai.net/agentic-langgraph-from-first-principles-to-a-working-multi-agent-system-9034d56a5b38 was originally published in Towards AI https://pub.towardsai.net on Medium, where people are continuing the conversation by highlighting and responding to this story.