aDNA Startup First Hour

One hour, four stages, one project-aware AI agent. For startup CTOs and founding engineers on a 10–50 person team: ship weekly, hire monthly, and stop institutional knowledge from leaking into Slack threads. By the end you have CLAUDE.md, STATE.md, and one ADR — agent-addressable governance. Skim the stages, then run the clock.

The four stages

Each stage names a time budget, an action, and the file it produces. Work in order; the file produced in one stage is the input the next reads.

Minutes 0–10 · Clone and install

Fork the base template into your repo root. You get the triad (what/, how/, who/), the 14-entity base ontology, and a working CLAUDE.md scaffold.

Minutes 10–25 · CLAUDE.md

Edit the root CLAUDE.md to name your project, voice, and standing orders. This is the first file every agent session reads.

Minutes 25–40 · STATE.md

Write a one-paragraph operational snapshot: current phase, blockers, next steps. Every new hire and every new agent session reads this second.

Minutes 40–60 · First ADR

Capture one decision that currently lives in a Slack thread as what/decisions/adr_001_<slug>.md. Future agents and future hires read this instead of re-asking.

Minutes 0–10 · Clone and install

Fork the base template (Agentic-DNA) into your repo root. No code changes — get the scaffold on disk.

  • Produced: base CLAUDE.md, MANIFEST.md, and the what/ · how/ · who/ triad.
  • Checkpoint: ls what/ how/ who/ shows three populated directories.

Minutes 10–25 · Make CLAUDE.md yours

Open CLAUDE.md and edit three things: project name and one-line description in the header, voice and tone, and the standing orders your team cares about (merge-freeze windows, review policies, default branch). The Your First CLAUDE.md tutorial is the long-form version.

  • Produced: a customized CLAUDE.md naming your project and your agent-work norms.
  • Checkpoint: a fresh agent session opens with your project name and voice, not the generic template.

Minutes 25–40 · Seed STATE.md

Write a one-paragraph operational snapshot: "We are in Q2 of $PROJECT. Current focus: $FEATURE. Blockers: none / <blocker>. Next up: <the 2–3 things that matter this week>." Keep it under 30 lines. The governance files concept page explains why STATE.md is the "what's happening now" slot.

  • Produced: STATE.md with current phase, blockers, and next steps.
  • Checkpoint: a new hire or agent session reads CLAUDE.md + STATE.md and answers "what is this project and what are we doing this week" without asking.

Minutes 40–60 · File the first ADR

Pick one decision currently living only in Slack. Create what/decisions/adr_001_<slug>.md with three sections: Context, Decision, Consequences. Two paragraphs per section is plenty. The knowledge-graph concept page explains how ADRs cross-link into the vault.

  • Produced: what/decisions/adr_001_*.md — the seed of your decision log.
  • Checkpoint: git commit, push, open a fresh agent session, ask "why did we choose <X>?" The agent finds the ADR without being pointed at it.

End of hour · What changed

Three files that didn't exist at minute zero now move the project from "ambient knowledge" to "agent-addressable knowledge."

Vault is legible to an agent

CLAUDE.md + STATE.md means a fresh agent session reads two files and knows where everything is. No per-session re-explanation.

Decisions stop evaporating

One ADR seeds a decision log. Within a month the founding-engineer bus factor starts breaking down — the "why" lives on disk.

Onboarding has an entry point

Your next hire reads STATE.md and walks the ADR list. No Slack archaeology, no "ping me if you have questions."

Self-reference: this vault ran this hour

The aDNA.aDNA vault went through its own version on 2026-04-13 — a forked base template, a customized CLAUDE.md, an initial STATE.md, and a first ADR under what/decisions/. The bootstrap session log is in how/sessions/history/; browse it to see the same hour play out on a real vault.

Hour two and beyond

Don't scale prematurely. Run the vault for a week on three files before adding more structure. A second decision earns adr_002; work bigger than a session graduates to a mission file. The startup persona page sketches the 3–6 month trajectory.

Next Steps

Tutorial: Your First CLAUDE.md

The long-form tutorial for the 10–25 minute stage. Go deeper when you have 20 minutes to invest in the agent-orientation file.

Governance Files

The four root-level files (CLAUDE.md, MANIFEST.md, STATE.md, AGENTS.md) and what each one does. Reference this when you graduate past hour one.

Startup Persona

Adopter profile: pain points, typical ontology extensions (api_spec, onboarding_checklist, customer_feedback), and deployment pattern.

Tutorial: Design a Mission

Hour one is individual discipline; hour two is team coordination. When you have a sprint-sized goal, a mission file turns it into claimable objectives.