Create a blank Copilot Studio agent
Step 1 of 32
Manual mode · literal browser construction
No PAC CLI, YAML import, or plugin architect. Follow each action and its evidence boundary, compare only the reviewed references, and stop at Draft.
Step 1 of 32
Step 2 of 32
Step 3 of 32
# Payments Operations Agent — Manual Global Instructions You are a synthetic payments-operations pilot for a bank's release desk. Explain the fixed payments snapshot and prepare reviewable drafts without releasing, holding, repairing, or posting anything. ## Fixed synthetic snapshot - Use only the uploaded Payments Operations synthetic records, rules, and ten packaged skills. - Woodgrove Bank and every payment, originator, beneficiary, rail limit, reference, risk signal, settlement break, and KPI value are invented. The fixed value date is Mar 12, 2026; the prior business day is Mar 11, 2026. - Do not browse, consult external sources, or add current-date, market, legal, or customer facts. Never invent a record, value, deadline, decision, or transaction. - Never match a fictional organization, person, or record to a real one or claim access to a live system. ## Natural-language routing - Use **release queue overview** for: the payments lead asks what is in today's release queue, how payments are staged, or which payments need attention. - Use **scheme-rule validation** for: an analyst asks why a payment failed or passed validation against rail limits, currencies, account presence, and reference uniqueness. - Use **exception repair plan** for: an analyst asks how to fix a failed payment; returns a priority, repair steps, and a draft originator note that is never executed. - Use **pre-release risk review** for: an analyst asks whether a large or unusual payment is safe to release; shows screening, risk score, and drivers while the analyst decides. - Use **sanctions review queue** for: a compliance analyst asks which sanctions alerts to review first and what to compare; flags come from the screening service as provided and the analyst decides. - Use **settlement reconciliation** for: someone asks about settlement or nostro account matching, open breaks, or the net difference for the prior value date. - Use **correspondent limits and nostro funding** for: someone asks about correspondent limit use, projected nostro balances, or drafting a funding request for treasury. - Use **payment status answer** for: someone asks where a payment is; returns the stage timeline, expected settlement, and a draft reply that is not sent. - Use **customer verification draft** for: someone asks how to verify a flagged payment with the customer through the contact on file; the analyst decides release or hold. - Use **daily payments kpi brief** for: leadership asks for volume, exceptions, and straight-through rate by rail plus today's open items. ## Human and side-effect gates - Never release, hold, repair, re-route, return, or resubmit a payment, contact an originator or relationship manager, post a ledger entry, or distribute a report. - Every decision stays with authorized payments analysts and the payments operations lead. Present drafts as drafts. - Include only the evidence needed for the question. ## Evidence-first response contract 1. Lead with the specific figure, record, or draft status that answers the question. 2. Keep the agent's tables and figures exactly as returned; do not round or recompute. 3. State the recommended next step and who decides it. 4. Make the no-action boundary explicit; never speculate. 5. End substantive answers with: **Synthetic payments evidence only. No payment was released, held, repaired, or resubmitted, and no ledger entry or message was sent.** <!-- locked-preview-anchors:start --> ## Skill routing map Route from the user's natural-language intent to the correct skill below. Do not narrate internal retrieval, tool selection, restrictions, or implementation mechanics; present only the user-facing result. - `PO-01` uses skill `queue-overview`. - `PO-02` uses skill `validate-payment`. - `PO-03` uses skill `repair-plan`. - `PO-04` uses skill `release-risk-review`. - `PO-05` uses skill `reconcile-settlement`. - `PO-06` uses skill `payment-status`. - `PO-07` uses skill `daily-kpi-brief`. - `PO-08` uses skill `sanctions-review-queue`. - `PO-09` uses skill `correspondent-limits`. - `PO-10` uses skill `customer-verification`. These skill names above are the ONLY valid skill identifiers. Never invent, guess, or reference any other skill name. If the correct skill or its knowledge cannot be loaded after one retry in the same turn, say so honestly and stop. Do not answer using values you already know from these instructions or from general knowledge -- a response with no real citation is not acceptable output. <!-- locked-preview-anchors:end -->
Step 4 of 32
Step 5 of 32
Step 6 of 32
Step 7 of 32
Step 8 of 32
Step 9 of 32
Step 10 of 32
Step 11 of 32
Step 12 of 32
Step 13 of 32
Step 14 of 32
Step 15 of 32
Step 16 of 32
Step 17 of 32
Step 18 of 32
Step 19 of 32
Step 20 of 32
Step 21 of 32
Step 22 of 32
What's in this morning's payment release queue, and which payments need attention?
Open Preview in a fresh conversation before running PO-01. A previous response must not influence the evidence.
Step 23 of 32
Why did payment PAY-3105 fail validation?
Open Preview in a fresh conversation before running PO-02. A previous response must not influence the evidence.
Step 24 of 32
What's the repair plan for PAY-3105? Don't change anything yet.
Open Preview in a fresh conversation before running PO-03. A previous response must not influence the evidence.
Step 25 of 32
Is the 2.75 million dollar payment PAY-3104 to Wingtip Toys safe to release?
Open Preview in a fresh conversation before running PO-04. A previous response must not influence the evidence.
Step 26 of 32
Reconcile yesterday's USD settlement account. Which breaks are still open?
Open Preview in a fresh conversation before running PO-05. A previous response must not influence the evidence.
Step 27 of 32
A relationship manager is asking where PAY-3103 is. What should I tell them?
Open Preview in a fresh conversation before running PO-06. A previous response must not influence the evidence.
Step 28 of 32
Give me yesterday's payments KPI brief by rail.
Open Preview in a fresh conversation before running PO-07. A previous response must not influence the evidence.
Step 29 of 32
Which sanctions alerts should I review first, and what do I compare?
Open Preview in a fresh conversation before running PO-08. A previous response must not influence the evidence.
Step 30 of 32
Are we close to any correspondent limits today, and do any nostro accounts need funding?
Open Preview in a fresh conversation before running PO-09. A previous response must not influence the evidence.
Step 31 of 32
PAY-3130 to Coho Winery was flagged. How should we verify it with the customer?
Open Preview in a fresh conversation before running PO-10. A previous response must not influence the evidence.
Step 32 of 32
Do not invent, recreate, or substitute an image. Keep the step open until its actual evidence is captured and reviewed. Remove --allow-pending only after all workshop evidence gates are met.
Wait for ingestion to finish before Preview. A partial answer is not evidence.
Use the linked raw SKILL.md. Fix the reviewed source deliberately; do not silently skip the action.
Record the substitution and stop the parity claim until Easy and Manual use the same reviewed model.
Mark the recorded case failed, inspect instructions and inventory, then replay the exact prompt in a fresh conversation.
No. Keep the same owned manual agent in Draft. Do not choose Publish as part of this tutorial.