AI Skills For The Full Deal Lifecycle
The AI skills we use to run finance work end to end — from screening a sector to defending a recommendation at committee. Each one is a repeatable, step-by-step method an AI agent follows, so the output is consistent, reviewable, and built the way a deal team would build it.
- Skills
- 22
- Deal Stages
- 5
- Valuation Methods
- 4
The Business Deal Process, Solved With Agent Skills
A deal used to mean weeks of hand-built screens, models and memos passed between people. Every step of that process is now covered by an agent skill — a defined method that runs the same way every time and is checked by your team. Chain them together and the whole deal process is handled end to end, from the first sector screen to the committee decision.
- Step 1
Origination
Decide where to focus and who would transact.
3 agent skills
- Step 2
Modeling & Valuation
Build the model, value the asset, stress-test it.
9 agent skills
- Step 3
Diligence
Turn the data room into findings and risks.
4 agent skills
- Step 4
Sell-Side Process
Market the asset and run an orderly process.
3 agent skills
- Step 5
Investment Committee
Land a clean, defensible decision.
3 agent skills
Origination
Decide where to focus and who would transact.
Modeling & Valuation
Build the model, value the asset, stress-test it.
Diligence
Turn the data room into findings and risks.
Sell-Side Process
Market the asset and run an orderly process.
Investment Committee
Land a clean, defensible decision.
Build Your Own Deal Agents
Enter your email and we’ll send you all 22 skills as a single markdown file, straight to your inbox. Paste a skill into your AI agent’s instructions, or save it as a SKILL.md, and use it to build your own deal agents.
- Every skill’s steps, example prompts and expected output
- One
.mdfile, ready for any agent or LLM tool - Organised by deal stage so you can chain them into a workflow
Want These Skills Running On Your Deals?
We set up AI agents that follow these methods on your own models, data rooms, and templates — with your team reviewing every output.