How to Sketch a LangGraph StateGraph from an Agent Brief
StateGraphs invent extra critic nodes. Nodes from the agent steps. REDUCER NOT IN INPUTS unless a list key is named.
The matching generator is the LangGraph StateGraph Nodes Edges and Reducers from an Agent Brief (No Invented Tools) (Langgraph Stategraph Nodes Edges Reducers From Agent Brief No Invented Tools) prompt. Browse related cards in the PromptDig library (Browse more prompts). When a filled run survives, share the version you actually use (Share a prompt).
Lock LangGraph version or write UNKNOWN
Sketch a LangGraph StateGraph with nodes, edges, and reducers from an agent brief. Do not invent tools or node names not listed. Start by filling Inputs, not by asking the model to remember last week's run. If a field is blank, write NONE or NOT IN INPUTS and leave it blank through Generate. The card is built so the model cannot honestly invent a number, owner, URL, or command that you did not paste.
Paste these fields before you hit run:
Agent brief (goal, steps, tools allowed): [Brief]
LangGraph version I lock (or UNKNOWN): [Version]
Tools I allow (or NONE): [Tools]
State keys I allow (or UNKNOWN): [State]
Words I must not use: [Banned]
What I must never invent: [Never]
Output format: [Format]
Language: [Lang]
Model I lock (or UNKNOWN): [Model]
That inventory is the honesty ledger. Anything that does not appear there is forbidden in the draft. If you catch yourself adding a nice-to-have after the run, you are no longer using the card. You are ghostwriting. Put the extra fact in Inputs and run again.
Nodes one-for-one with brief steps
Generate is numbered on purpose. Do not skip a step because the first paragraph looked done. The early steps exist to stop later prose from smuggling claims.
Walk the Generate list in order:
- Honesty ledger: Brief nouns, Version, Tools, State, Model, Lang. Forbidden: extra tools, extra nodes, inventing checkpointers.
- Version line: quote Version or VERSION UNKNOWN.
- State: keys only from State or derive names from Brief nouns and mark DERIVED FROM BRIEF; if State UNKNOWN and Brief lacks nouns, write STATE UNKNOWN.
- Nodes: one per step in Brief. No extra nodes.
- Edges: linear unless Brief names a branch. Conditional edge only if Brief has a branch word.
- Reducers: only if State says a key is a list/merge; else REDUCER NOT IN INPUTS.
- Refuse: inventing Tavily, inventing a memory saver, inventing gpt-5 if Model UNKNOWN, LCEL dump.
- Compliance pass: Banned and Never. Format as Format.
If a step asks for a version lock, quote the version from Inputs in the output. If a step asks for a refuse list, keep the refuse list in the published artifact, not in a sidebar you delete. Reviewers should see what the model was not allowed to do.
Conditional edges only when the brief branches
Most failures are the same shape: a missing field gets a confident fill. A conversion rate appears. A Gradle task appears. A flash point appears. A caption appears on a job that asked for slide text only. Your review is to search the draft for numbers, names, and commands, then grep Inputs. No match means cut.
Honor the constraints as hard stops, not vibes:
- LangGraph StateGraph from an agent brief. Not a workflow-brief node-only card.
- Never invent tools, extra nodes, or model IDs.
- No emojis.
When the card says not legal advice, not certification, not an exam dump, or not a caption engine, that sentence belongs at the top of the output. Deleting it to look more finished is how you inherit risk.
Refuse invented tools and savers
Finish with the compliance pass the prompt already asks for. Quote the banned-word hits. Cut them. Print character counts when the job has a cap. Print word counts when the job has a budget. List gaps as gaps. Five missing facts are more useful than one smooth paragraph.
Tags on the card (langgraph stategraph nodes edges, agent brief reducers sketch, no invented langgraph tools) are a reminder of the job shape, not an invitation to wander into a neighboring cluster. If you need a different surface, open a different PromptDig card rather than stretching this one.
Fill the card, then run
Replace every bracket. Run on ChatGPT, Claude, or Gemini. Read the ledger first, then the artifact. If the model invents a commit, KPI, DOI, PEL, bid, or logo, discard the run. Tighten Inputs. Run again. Share the filled card that survived, not the first draft that sounded done.