How to Write OpenAI Assistants API System Instructions
Assistants instructions invent function names. Allowlist tools from the brief. MODEL UNKNOWN if blank.
The matching generator is the OpenAI Assistants API System Instructions from a Tools Brief (No Invented Tools) (Openai Assistants Api System Instructions From A Tools 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).
Quote the product as OpenAI Assistants API
Write Assistants API system instructions from a tools brief. Do not invent tools, function names, or model IDs not locked in Inputs. 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:
Tools brief (who it helps, allowed tools, what a good answer looks like): [Brief]
Model I lock (or UNKNOWN): [Model]
Tools I allow (file_search / code_interpreter / function names as pasted / NONE): [Tools]
Function signatures I allow (or NONE): [Functions]
Words I must not use: [Banned]
What I must never invent: [Never]
Output format: [Format]
Language: [Lang]
Product I lock: [Product or OpenAI Assistants API]
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.
Allowlist only the tools in Inputs
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, Model, Tools, Functions, Product, Lang. Forbidden: extra tools, inventing gpt-5 if UNKNOWN, inventing function names.
- Role block: one paragraph from Brief. Product quoted.
- Goal block: quote the goal in Brief. Do not add extra success metrics.
- Tools allowlist: only Tools. If a tool is missing, write NOT IN INPUTS. Function names only from Functions.
- Input/Output blocks from Brief only.
- Constraints: copy hard lines from Brief. Model quoted or MODEL UNKNOWN.
- Refuse: Custom GPT OpenAPI invent, Cursor .mdc, ChatGPT Projects starters, invented vector-store IDs.
- Compliance pass: quote Banned and Never hits. Cut them. 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.
Leave vector-store IDs blank
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:
- OpenAI Assistants API system instructions from Brief. Not a Custom GPT listing and not Cursor rules.
- Only Tools and Functions named in Inputs.
- Never invent model IDs or extra tools.
- 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 Custom GPT and Cursor packs
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 (openai assistants api instructions, tools brief system instructions, no invented assistant 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.