Skip to content
← All templates
🧭

Startup Decision Log

The why behind every meaningful choice your company has made. Strategy, product, people. Claude and ChatGPT read it the next time anyone asks "why did we go this way".

Requires an AI connected to your Hjarni account via MCP.

https://hjarni.com/templates/startup-decision-log

Copy this URL and paste it into Claude or ChatGPT to install the template.

How to use

  1. 1 Share this page. Paste this URL into Claude or ChatGPT. Your AI reads the template definition and installs it.
  2. 2 Folders, tags, and instructions appear. Your AI creates the full structure in your Hjarni account, ready to use.
  3. 3 Start adding notes. The AI instructions guide your AI on where to put things and how to organize them.

The reasoning behind every choice your company has made. Saved once. Searchable forever.

Most early-stage decisions live in someone's head, a Slack thread, or a deck no one opens twice. Six months later, the team reopens the same argument from scratch. This template gives those decisions a shape your AI can actually read.

A decision note format

Every decision follows the same five sections. The starter note in each folder is exactly this skeleton.

# Decision: <Short noun phrase>

Context:
Options considered:
Decision:
Tradeoffs accepted:
Date:

Why the Superseded folder matters

When a new decision replaces an old one, the old reasoning is still useful. Knowing what you used to believe is part of the answer to "why did we change". The Superseded folder keeps the trail intact instead of letting old context disappear quietly.

A workflow that earns the template's keep

  1. When the team makes a real decision, save it in the matching folder.
  2. When asked "why did we pick X", ask Claude or ChatGPT instead of digging through Slack.
  3. When a new call overrides an old one, move the old note to Superseded and link the replacement.

A new hire connects their AI. They ask "why does pricing work this way". They get the real story, with tradeoffs and date, the first time they ask.

A real example

Eighteen months in, your second engineer is making a build vs buy call on background jobs. They ask Claude, "How have we made past build vs buy decisions?" Claude reads Product, returns three past decisions with their tradeoffs, and explains the pattern. The new call inherits the reasoning, not just the outcome.

Common questions

Common questions

How is this different from a meeting notes folder?

Meeting notes capture what was said. This template captures what was decided, with the reasoning intact. The two work well together.

Do I have to use the exact five sections?

The folder instructions ask Claude or ChatGPT to follow Context, Options, Decision, Tradeoffs, Date. You can edit per folder, but the consistency is what makes the AI's answers reliable.

Can my team share this?

Yes. Install in a team space so every teammate's Claude or ChatGPT reads from the same decision log.

Will the AI invent decisions I never made?

Folder-level instructions tell each AI to refuse to fabricate decisions. If a decision is not in the folder, the AI says so plainly.

Related pages

Templates overview | Developer Project Memory template | Meeting Memory template | Hjarni for Founders | Hjarni for Teams | Custom AI instructions per folder

Structure

Tags

strategy product people superseded tradeoff

Folders

Startup Decision Log
Every decision that shaped the company. Context, options, tradeoffs, and date. Searchable forever.
Has AI instructions
5 starter notes
Strategy
Market, pricing, positioning, fundraising, who you serve.
Has AI instructions
Product
Build vs buy, scope cuts, technical bets, feature kills.
Has AI instructions
People
Hiring bets, role designs, comp bands, structure changes.
Has AI instructions
Superseded
Old decisions that a newer call overrides. Kept so the trail survives.
Has AI instructions

For your AI

Share this page with your AI. It reads the definition below, creates the folders, instructions, tags, and starter notes in your account.

Show template definition

Install steps for AI agents:

  1. Check existing tags with tags-list. Only create missing ones with tags-create.
  2. Create containers top-down using containers-create, noting the returned IDs. Use parent_id to build the hierarchy.
  3. For each container with llm_instructions, call instructions-update with level: "container" and the container's ID.
  4. Create any seed notes using notes-create, placing them in the correct container by ID. Use container_path to resolve which container.
  5. Discuss any customizations with the user before or after installing.
---
name: Startup Decision Log
description: 'The why behind every meaningful choice your company has made. Strategy,
  product, people. Claude and ChatGPT read it the next time anyone asks "why did we
  go this way".

  '
tags:
- strategy
- product
- people
- superseded
- tradeoff
containers:
- name: Startup Decision Log
  description: Every decision that shaped the company. Context, options, tradeoffs,
    and date. Searchable forever.
  llm_instructions: |
    This is a startup decision log. The point is to make the reasoning behind past calls retrievable, so the team and the AI both stop reinventing the same arguments.
    - When asked "why did we pick X" or "why did we drop Y", search this folder first and quote the matching decision note.
    - Never invent a decision rationale. If the answer is not here, say so plainly and ask the user.
    - Treat each decision as a snapshot at a point in time. Do not silently edit past decisions. If a new decision overrides an old one, create a new note and move the old one to Superseded.
    - When the user resolves something in conversation that looks like a real decision, suggest saving it here in the standard structure.
    - Each decision note should follow: Context, Options considered, Decision, Tradeoffs accepted, Date.
  children:
  - name: Strategy
    description: Market, pricing, positioning, fundraising, who you serve.
    llm_instructions: |
      Use this folder for company-level strategic calls.
      - When asked about pricing changes, ICP shifts, fundraising terms, or positioning bets, search here first.
      - Tag every note with "strategy".
      - Keep one decision per note. If a single discussion produced three real decisions, write three notes.
  - name: Product
    description: Build vs buy, scope cuts, technical bets, feature kills.
    llm_instructions: |
      Use this folder for product and engineering bets.
      - When asked about build vs buy, technology choices, deprecations, or feature cuts, search here first.
      - Tag every note with "product".
      - Link to relevant architecture or runbook notes when the decision has operational fallout.
  - name: People
    description: Hiring bets, role designs, comp bands, structure changes.
    llm_instructions: |
      Use this folder for decisions about the team.
      - When asked about reporting structure, comp philosophy, role design, or hiring bets, search here first.
      - Tag every note with "people".
      - Be discreet. Reference roles and decisions, not personal evaluations. Keep individual performance notes elsewhere.
  - name: Superseded
    description: Old decisions that a newer call overrides. Kept so the trail survives.
    llm_instructions: |
      Use this folder for decisions that have been replaced by a newer call.
      - Never delete an old decision when it is replaced. Move it here and link the replacement.
      - When asked "did we used to do X", search this folder.
      - Each note in here should link forward to the decision that superseded it.
  notes:
  - title: 'Decision: Charge per seat, not per workspace'
    body: |
      A starter decision note. Replace with a real one.

      ## Context
      Early customers were sharing one workspace across whole teams. Revenue did not grow with usage.

      ## Options considered
      1. Stay on workspace pricing and raise the base price.
      2. Move to per-seat pricing with a free first seat.
      3. Usage-based pricing tied to notes stored.

      ## Decision
      Per-seat with a free first seat. Solo users stay free. Teams pay for additional members.

      ## Tradeoffs accepted
      Slightly more friction for solo expansion. In exchange, revenue tracks real usage and team buyers have a clearer mental model.

      ## Date
      Replace with the real date when you adopt this template.
    tags:
    - strategy
    - tradeoff
    container_path: Startup Decision Log > Strategy
  - title: 'Decision: Build the MCP server ourselves'
    body: |
      A starter product decision. Replace with a real one.

      ## Context
      We needed a clean integration with Claude and ChatGPT. The market had no off-the-shelf option that matched our model.

      ## Options considered
      1. Wait for a third-party MCP gateway to mature.
      2. Build the server in-house alongside the app.
      3. Ship an export-only integration and skip live access.

      ## Decision
      Build in-house, owned by the core team.

      ## Tradeoffs accepted
      More surface area to maintain. In exchange, full control of the tool surface and faster iteration.

      ## Date
      Replace with the real date when you adopt this template.
    tags:
    - product
    container_path: Startup Decision Log > Product
  - title: 'Decision: One generalist hire before a second engineer'
    body: |
      A starter people decision. Replace with a real one.

      ## Context
      Two open hires, one engineer one generalist. Funding allowed for one immediately.

      ## Options considered
      1. Hire the second engineer first to ship faster.
      2. Hire a generalist to own marketing, support, and operations.

      ## Decision
      Hire the generalist first.

      ## Tradeoffs accepted
      Engineering velocity stays where it is for the next quarter. In exchange, distribution and customer feedback loops get an owner.

      ## Date
      Replace with the real date when you adopt this template.
    tags:
    - people
    container_path: Startup Decision Log > People
  - title: 'Superseded: Free unlimited workspaces'
    body: |
      A starter superseded note. Replace with a real one.

      ## Original decision
      Workspaces were free and unlimited for all users.

      ## Why it was replaced
      Per-seat pricing replaced this approach. See [[Decision: Charge per seat, not per workspace]].

      ## Date superseded
      Replace with the real date when you adopt this template.
    tags:
    - superseded
    container_path: Startup Decision Log > Superseded
  - title: 'Decision: Drop the Slack integration roadmap item'
    body: |
      A starter product decision about killing scope. Replace with a real one.

      ## Context
      Slack integration was on the roadmap for Q2. Two customer interviews suggested it was not the bottleneck.

      ## Options considered
      1. Build a minimal Slack integration anyway.
      2. Drop it and reallocate the time to mobile.
      3. Defer and revisit in Q4.

      ## Decision
      Drop and reallocate to mobile.

      ## Tradeoffs accepted
      We lose a talking point with Slack-heavy buyers. In exchange, mobile ships a quarter earlier.

      ## Date
      Replace with the real date when you adopt this template.
    tags:
    - product
    - tradeoff
    container_path: Startup Decision Log > Product

Write once. You both remember.

Free to start. No credit card required.

Give your AI a memory

Works with Claude and ChatGPT today. Gemini coming soon.