Template Library
Models, decks, and cheatsheets you can fork and adapt in minutes.
Armory · Tools & Resources
Stop rebuilding from scratch. Deploy from the Armory.
Reusable, version-controlled assets for real deal work: models, decks, cheatsheets, battle-tested prompts, and agent packs that draft comps, precedents, and memos. Search, fork, deploy, ship.
Why it matters
Your scarcest asset is hours. You are burning them rebuilding what should already exist.
Scenario automation alone saves finance teams 3 to 5 hours a week, and AI around the edges saves another 2 to 4. Across a year that is roughly 200 hours per analyst pulled back from pure document review, with research throughput rising 3x to 5x. The Armory hands you those workflows pre-built so you capture the upside without engineering it yourself.
Starting from a vetted asset raises your floor. Open a sound template instead of a blank workbook and the mechanical scaffolding is already correct, so your attention goes straight to the judgment calls that move the deal. The leading AI still underperforms a strong junior on quality and judgment, so human review stays mandatory. The Armory is built around that reality.
Source: Knowledge / AI Capability Assessment 2026 · Perplexity research, 2026
/Get hours back, every week.
Scenario automation alone saves finance teams 3 to 5 hours a week, and the Armory hands you that workflow pre-built.
/Ship to a higher floor.
Start from a vetted template rather than a blank workbook, then spend your time on the judgment calls that actually move the deal.
/Prompts that survive a live model.
Battle-tested IB prompts for comps, precedents, and IC memos, refined on real tasks rather than demo screenshots.
/Version-controlled, so it compounds.
Fork an asset, improve it, share it back. Your toolkit gets sharper with every deal.
What's inside
A working arsenal, not a static download page.
Every pack ships with its guardrails attached. The known failure modes, confident hallucination of figures, circular-reference breakage, GAAP or IFRS misapplication, and audit-trail gaps, are flagged inside the asset itself, so you deploy with your eyes open.
Prompt Vault
Battle-tested prompts for the IB tasks you run constantly, refined on real work so they hold up against a live model.
Agent Packs
n8n and Claude workflows for comps, precedents, and memos, built around Claude for Excel reading the entire workbook before it responds.
Data Feeds
Free and affordable APIs and scrapers so you spend less time hunting for inputs. Built on n8n, Obsidian, and Claude.
Proof and credibility
Claude for Excel reads the entire workbook, every tab, formula, and dependency, before responding, and a full 11-tab DCF can be drafted in about 10 minutes; the Agent Packs are built around that. We ship the catch, not just the speed: the same models hallucinate confidently, which is exactly why every pack states its failure modes and data-handling assumptions up front.
Source: Knowledge / AI Capability Assessment 2026
Straight answers
Before you ask.
I already have my own templates, why pay?
The Armory is maintained and version-controlled against current standards and current AI tooling, and the Agent Packs do what a folder of old .xlsx files cannot. Your static templates do not update themselves; ours do.
Are the prompts actually any good?
Yes. They are battle-tested on real IB tasks, and each ships with the failure modes it is designed to avoid, so you know its limits before you trust it.
Is it safe to run agents on deal data?
We teach the boundary plainly: use AI for first drafts and structure, never raw MNPI through public tools without the right controls. Every pack states its data-handling assumptions.
Will this make me lazy or deskill me?
Only if you let the AI think for you. Every asset accelerates execution while keeping the judgment in your hands: a force-multiplier on real fundamentals, never a shortcut around them.
