Standalone skill pack for turning existing diligence materials into a structured deal, IC, partnership, or acquisition memo.
This skill helps an AI client draft a real decision memo from work that already exists. It takes market findings, model review output, meeting synthesis, and supporting documents, then shapes them into a recommendation-ready memo while keeping evidence gaps and confidence levels explicit.
For the OB1 composition workflow that feeds this skill from earlier analysis steps, use the Research-to-Decision Workflow recipe.
- Claude Code
- Codex
- Cursor
- Other AI clients that support reusable prompt packs, rules, or custom instructions
- Working Open Brain setup if you want the skill to use memory search or capture (guide)
- AI client that supports reusable skills, rules, or custom instructions
- A real diligence packet, not just a company name and a blank page
- Copy
SKILL.mdinto the reusable-instructions location for your AI client. - Restart or reload the client so it picks up the skill.
- Test it with a prompt like:
Draft an IC memo from this research summary, model review, and meeting notes.
For Claude Code, a common install path is:
mkdir -p ~/.claude/skills/deal-memo-drafting
cp skills/deal-memo-drafting/SKILL.md ~/.claude/skills/deal-memo-drafting/SKILL.md- "Draft a deal memo"
- "Write the IC memo"
- "Turn this diligence packet into a memo"
- "Draft the partnership recommendation"
- "Write the acquisition brief"
When installed and invoked correctly, the skill should produce:
- a structured memo with thesis, market, business, economics, risks, and recommendation
- explicit open questions instead of hidden assumptions
- confidence signaling that matches the evidence quality
- optional Open Brain capture of the final memo summary or recommendation
Issue: The memo sounds polished but thin Solution: Provide the actual underlying diligence materials. This skill is strongest when it has research, model review, and meeting outputs to work from.
Issue: The draft hides uncertainty Solution: Preserve the rule that evidence gaps stay visible. The memo should never become more certain than the diligence allows.
Issue: The skill gets used too early
Solution: If the source set is still messy or contradictory, run research-synthesis first. If the economics are still unreviewed, run financial-model-review first.
Keep the core behavior intact even if your client uses different tool names. The skill should draft from diligence, separate fact from inference, and finish with a recommendation that reflects the evidence instead of smoothing over it.