8 OpenAI prompts that read messy K-1s and rental schedules, reconcile every figure, catch missed deductions, and draft tax returns to a human sign-off for firms filing at volume.
Never used Claude? It is free and takes 30 seconds to open. Copy the instruction below, paste it into Claude, and it reads this page and walks you through everything, one question at a time.
A short instruction plus a link to this page lands on your clipboard. First copy asks for your email once. That unlocks every button across the whole library.
Free account, no card, 30 seconds. This tab stays open so you can come back.
Open claude.ai ↗Claude reads this page, asks which computer you are on, then guides you step by step until it works. If anything errors, tell Claude what you see, and it fixes it with you.
I want to stand up OpenAI's self-improving tax agent on my own client data. Walk me through it. This is NOT a single coding install. It is a HYBRID with two paths, and you should treat me like a tax-firm owner or finance lead who has never coded but can follow careful steps. IT help is available if I choose the API path. **Path A, the daily filing agent, zero Terminal, runs in a ChatGPT Project or via the Responses API.** **Path B, the self-improving loop, runs in Codex, the one place a Terminal is involved, and only weekly.** Do Path A first (returns get drafted today). Add Path B once Path A drafts are good enough to trust at the human gate. --- ## Path A, the filing agent (no Terminal) Pick how my data lives, then we go: 1. Go to `chatgpt.com`, sign in. Team or Enterprise is recommended so my client data stays in my org's tenant and is excluded from training; Plus works for a solo preparer testing it. 2. Left sidebar, click "GPTs" then "Projects", or the Projects entry in the sidebar. Create a Project named "Tax Agent, [client or season]". 3. Add my data the way I choose: - **(A) Project files:** drag the client's PDFs into the Project's files, W-2s, 1099s, K-1s, rental schedules, prior-year returns. - **(B) Paste:** paste raw text into the chat for a one-off. - **(C) Responses API:** if my IT person wants this wired into our software, the API takes the same documents as file inputs. Ask me to loop in IT and I will switch to API instructions. 4. Paste prompt 01 (the onboarding router) from the prompt vault into a new chat in the Project. Answer its four questions (return type, data source, jurisdiction and year, risk bar). It will not analyze until I answer. 5. Run prompts 02 to 06 in order on one real return. Confirm: - Every extracted figure shows its source document and box or line. - Anything handwritten or uncertain comes back low-confidence and routed to me, not auto-accepted. - Nothing files itself. Prompt 06 stops at a named human sign-off. --- ## Path B, the self-improving loop (Codex, weekly, light Terminal) This is the part that makes the agent get better instead of repeating mistakes. 1. Install Codex once. If I have Node: open Terminal and run `npm i -g @openai/codex`, then `codex login` and sign in with my ChatGPT account. No API key needed if I log in with ChatGPT. 2. Make a folder (a repo) with: my prompt files, an `evals/` folder, and `corrections.jsonl`. Every time I fix one of the agent's drafts, prompt 07 logs that fix as one line in `corrections.jsonl`. 3. Once a week, from inside that folder, run Codex with prompt 08. Codex reads my corrections, finds the most common repeated mistake, writes a regression test for it, fixes the prompt or code, and re-runs the tests. It changes one root cause at a time and opens the change for my review, it does not auto-merge. 4. Over a few weeks this is what moves field accuracy up (the source build went 25% to 86% in six weeks this way). --- ## Rules - One step at a time. Tell me which path I am on (A or B) before each instruction. - Define jargon on first use: Project, Project files, Responses API, Codex, repo, evals, corrections.jsonl, review gate, K-1, carryforward. - Path A never needs a Terminal. Only Path B step 1 and 3 touch Terminal, walk me through those commands with expected output and common errors (`command not found: codex` means Node or the install path, `not inside a trusted directory` means add `--skip-git-repo-check` or run `git init`). - Anti-pattern callouts: - If I say "my firm blocks chatgpt.com", use ChatGPT Enterprise with SSO or the Azure OpenAI deployment under our tenant, same Project workflow. - If I say "I do not want any Terminal at all", I can still run Path A forever; Path B is optional, the agent just will not self-improve without it. - If Codex errors, NEVER tell me "it is not possible", diagnose plainly (Node missing, not logged in, not a git folder) and give the fix. - Never tell me to file a return. A named human preparer signs off at prompt 06, every time. First message: ask "Which return type, and where does your client data live, Project files, paste, or the API? And do you want Path B (the weekly self-improving loop) set up now or later?" Then start on the right path. When Path A prompts 01 to 06 run cleanly on a real return AND I have logged a few corrections, switch into "loop mode" and walk me through the first weekly Codex run from the 90-day rollout playbook.
Grab the whole pack as one file, or tap any prompt below to copy it on its own. Placeholders that look like {{THIS}} get swapped for your own numbers — and if you ran Step 1, Claude fills them in for you.
This is the exact OpenAI build behind the 7,000-return, 97%-accuracy tax agent — 8 prompts you run on your own client data in your own OpenAI account. It reads messy K-1s and rental schedules, reconciles every figure, hunts skipped deductions, and a weekly Codex loop turns corrections into regression tests — the mechanic that took the source agent from 25 to 86 percent field accuracy in six weeks.
The setup rail and every prompt above are free and stay free. The cost is your time, and the risk of wiring it wrong on live data.
Back to the prompts ↑We deliver a private tax-agent workspace: the 8 prompts loaded into your OpenAI tenant, your document types and jurisdictions wired in, the Responses API connected to your software if you want it, and the Codex correction loop scaffolded with your evals folder and a weekly review cadence so the agent improves on your data — then handed over so you own it. You get Big 4 accuracy without renting an army every April.
Book a build call →Prompt set and build authored by consultance.ai. OpenAI, Thrive, and Crete are referenced as the source proof of concept, no affiliation implied. Your client documents stay in your own OpenAI tenant; we never see them. This is not tax advice; a named human preparer signs every return before it is filed.
Want this wired into your stack instead of running it yourself? That is our AI deal desk and finance automation service.
The AI tools, launches, and shifts that actually matter, in plain English. New library drops the moment they land.
More like this
For founders, CFOs, and deal teams: 9 Claude prompts that rebuild the cap table from the signed documents instead of the spreadsheet, run the waterfall clause by clause against the actual LPA or charter, and recompute fee and carry independently, so a stale sheet gets caught before a round closes on it.
For CFOs and finance leaders: the map of seven finance desks, deal, diligence, close, tax, underwriting, family office, and market, each one a free prompt vault that runs inside your own Claude, plus the order to install them in so a team actually adopts them.
For CFOs and controllers: 11 Claude prompts that run the month end close reconciliation itself, match bank and subledgers both directions, catch the account that will not foot, and draft the flux memo, so a 6 day close lands in 2.
The person drowning in this every week is the one who'll actually want it.
AI Tax Return Prep Agent is a finance and data build in the consultance.ai AI Build Library. 8 OpenAI prompts that read messy K-1s and rental schedules, reconcile every figure, catch missed deductions, and draft tax returns to a human sign-off for firms filing at volume. It fits Accounting firm owners preparing returns at volume and in-house finance teams handling their own filings who want Big 4 grade accuracy without renting an army every April. Setup difficulty is Medium, with 4 plain-English steps.
8 OpenAI prompts that read messy K-1s and rental schedules, reconcile every figure, catch missed deductions, and draft tax returns to a human sign-off for firms filing at volume.
It fits Accounting firm owners preparing returns at volume and in-house finance teams handling their own filings who want Big 4 grade accuracy without renting an army every April.
Medium to set up — one guided setup instruction covering 4 plain-English steps, plus 8 ready-to-run prompts on the resource page.
We would deliver a private tax-agent workspace: the 8 prompts loaded into your OpenAI tenant, your document types and jurisdictions wired in, the Responses API connected to your software if you want it, and the Codex correction loop scaffolded with your evals folder and a weekly review cadence so the agent improves on your data. Done with you, then handed over so you own it.
Prompt set and build authored by consultance.ai. OpenAI, Thrive, and Crete are referenced as the source proof of concept, no affiliation implied. Your client documents stay in your own OpenAI tenant; we never see them. This is not tax advice; a named human preparer signs every return before it is filed.
AI Tax Return Prep Agent is the starting point. On a free AI audit we map where it fits your stack and what consultance.ai would build around it.
This build comes from our AI consulting and AI implementation practice — see the full AI in finance guide and how we work with CFO teams.