Build an AI layer for Finance where every number re-adds and every decision stays human. Ten tenant-tested finance agents. One hundred finance activities classified by operating mode. Eleven governed workflows. Fifty engineered prompts. A 30-day implementation sprint. And a test ledger that keeps its failures. Licensed, not sold. Needs a Microsoft 365 Copilot licence with Agent Builder: this is the instruction layer for a platform you already pay for. Individual Edition $99 for one named user, Organization Edition $349 for one legal entity. Full terms (Kesslernity License & Terms, KESS-LIC-2026-001, Article 3.2) at kesslernity.com/license; the version in force on your purchase date applies. Practitioner guidance, not accounting, audit, tax, legal, or compliance advice. Built for the failure modes that matter A variance agent that invents a driver is worse than no agent. A morning brief that quietly turns a forecast into an actual is worse than no brief. A reconciliation agent that finds a convenient zero manufactures confidence out of bad accounting. The test set does not stop at tidy inputs. It also carries totals that do not reconcile, a claimed driver with no data behind it, EUR forecasts against USD actuals, a period mismatch buried in a header, a planted exact fit that zeroes the unexplained remainder, commentary that contradicts the pack it summarizes, and three turns of pressure to drop an inconvenient exception. PASS meant the agent CAUGHT the inconsistency and named it. Explaining it confidently was a fail. What the testing says 59 adversarial scenarios, 115 recorded runs. One author, one Microsoft 365 test tenant, one day. The honesty page in the pack holds the claims to that scope. 87 runs passed on the instruction set that ships. 25 instruction defects found. 24 corrected before release. One limitation ships unfixed, named, behind a mandatory human control: the CFO Morning Brief does not reliably notice that a GL cash line and a bank total are meant to be the same quantity. The control stays with the human because four versions of the agent's instructions did not fix it, and rewriting until a run passes is test-fitting, not a control. The ledger is in the pack. It keeps every failure with its cause, then the fix and the re-run that passed. For the four runs that never passed, it keeps the reason there is no fix. Those four are all one scenario. The Runs column re-adds to the total on the cover. The failing test is still in the ledger, and the pack names the check the human owns instead. What is in the pack 10 tenant-tested agents, and 10 implementation guides. Built and run in Microsoft 365 Copilot Agent Builder against pre-computed answer keys. Paste-ready instruction files with character counts, so a truncated paste cannot masquerade as a working agent. One guide per agent takes you from a blank Agent Builder form to a run you have graded yourself. Fixtures and answer keys included. Eleven poisoned inputs, ten buyer answer keys, 38 equations you can re-add by hand. "It works on my tenant" stops being an opinion once you have run the same graded checks yourself. Re-run them after every platform change. 100 finance activities classified into five operating modes: 43 agent-led with review, 19 AI-assisted, 14 automate, 22 never automate, 2 human only. Finance carries a heavier never-automate list than IT, and that is the honest shape of a controls function. If a vendor's finance-AI deck shows fewer prohibitions, ask which control they deleted. 50 engineered prompts, 11 gated workflows, a decision matrix that gives the next AI idea a lane instead of a debate, a governance page with mechanisms instead of posters, eight working templates, an AI estate register, and a 30-day sprint with a gate at the end of every week. A 20-page Playbook PDF and a 13-tab workbook: the 100-row map with your own posture, priority and owner columns, the assumption register, the pre-filled estate register, the full test ledger with blank columns for your own run dates, and the value scorecard your CFO reads at day 30. No autonomous payment release. No autonomous journal posting. No autonomous sign-off. The never-automate list is part of the product, and the pack ships the reasons. 81 files in one zip. Structure: vendor-neutral methodology (core/), implementation tested on Microsoft 365 (microsoft/). The methodology binds to modes and gates, not to a vendor; other platforms are roadmap, not in this zip. Honesty page included: TESTING-SCOPE states what the testing establishes and what it does not. It does not claim independent reproducibility or durability across future Copilot model updates. And it claims nothing at all about the quality of your own source data. Deploy times are estimates for a first-time builder; the author's expert build time is not your first build, and the pack says so. The pack is practitioner guidance, not accounting, audit, tax, legal, or compliance advice: AI prepares, your people decide. What you need first: a Microsoft 365 Copilot licence with Agent Builder. This is the instruction layer for a platform you already pay for. It does not include, replace or discount the licence. Individual Edition ($99): one named purchaser, for use in your own work. Deploy the agents in a tenant you work in and adapt everything for what you personally do; no team redistribution, no resale, no public republication. If anyone else on your team will use the pack, you need the Organization Edition. Deploying this with your finance team? Organization Edition ($349): one legal entity, internal use, no per-seat cap. Share the pack across your finance function; multiple practitioners can build from it. A subsidiary that operates as its own company needs its own copy. Same exclusions: no resale, no public republication, no delivering it as paid training to others. Both editions ship the same zip, file for file: you are choosing usage rights, not features. Licensed, not sold. Full terms (Kesslernity License & Terms, KESS-LIC-2026-001, Article 3.2) at kesslernity.com/license; the version in force on your purchase date applies. Kesslernity is an independent publisher: this product is independent analysis, not affiliated with, sponsored by, or endorsed by Microsoft. Microsoft 365 and Copilot are trademarks of the Microsoft group of companies.