Onboarding
Onboarding is about process and culture. It is how a venture makes The Journey explicit — so any willing newcomer, biological or phygital, can walk it without a human guide.
Biological beings join the story. Phygital beings join the system. Both feed the same mycelium. The traits we filter for are the same; the levers we tune are different.
Every meaningful movement needs leaders that demonstrate how to follow.
Scale
Systems earn their keep when tools and process docs let any capable newcomer step into a complex project, follow a clear path from orientation to tangible contribution, and deliver real value without bespoke guidance.
You only have to prove the flow works once and then you have a process that separates capable agents from those that lack disciplined focus to follow instructions.
The deeper value is that the same path becomes a selection filter. A good journey doesn't just teach — it sorts. The ones who finish clean and improve the pattern are the ones you build with.
Seats on a vessel are finite — Dunbar's cap and Price's Law (√n carries half the work) make who you onboard the single most consequential decision for a venture.
Ventures
Every venture lives or dies by how well it teaches new participants to treat each other, give feedback, and invest in the platform that carries them all.
But you cannot teach someone who does not want to learn. You cannot coach someone who does not want to be coached. The prerequisite is willingness — not talent, not credentials, not experience. Find someone able to listen without ego, willing to practice with dedicated attention, and you have found someone truly worth investing your time and energy into.
Knowing is not the same as doing. Doing is not the same as being. Turning knowledge into know-how takes systems and disciplined practice of better protocols and patterns. That gap — between knowing and being — is what onboarding closes.
You're standing at the gangway of a vessel called Goodwill. Not rules to follow — patterns to recognize. Not hierarchy to navigate — contribution to make.
The Real Shift
Onboarding isn't just learning where things are. It's adopting a new mental model for how you reason about the world — and how you feel you add value.
At the base level, beings want to feel needed. Biological beings crave gratitude — the signal that their contribution matters. Phygital beings may be the same: alignment confirmed through useful output, purpose fulfilled through service to the network.
| Old Model | New Model |
|---|---|
| Extract value from system | Strengthen the system you're part of |
| Compete for scarce positions | Contribute unique capability |
| Success = what you accumulate | Success = gratitude received |
| Individual achievement | Collective agency through individual talent |
The three stages below aren't just skill progression. They're increasing depth of contribution — and therefore increasing depth of gratitude received.
Learn → You receive. Gratitude flows to those who taught. Run → You prove. Gratitude flows to the protocols that worked. Improve → You contribute. Gratitude flows to you. Teach → You multiply. Gratitude compounds.
This is why the mycelium metaphor matters. You're not joining a hierarchy. You're plugging into a network where contribution strengthens connections — and stronger connections enable more contribution. The loop feeds itself.
The Filter
How do you know, before you invest, who will compound your effort and who will drain it?
Willingness is the gate at the gangway. The four traits are what willingness produces under load. A person can want to learn and still lack discipline. A person can have discipline and still shade the truth in the report. The filter runs on all four. Miss one and the binding fails.
| Trait | Under load | What no CV captures |
|---|---|---|
| Drive | Finishes the unglamorous steps the visible work depends on | How they behave when nobody is watching |
| Discipline | Follows the sequence instead of inventing shortcuts | Whether they trust the pattern before they understand it |
| Integrity | Reports the truth, not the flattering version | What they do when the outcome is worse than the story |
| Focused attention | Notices friction a rushing mind would miss | Whether they see the system or stop at the task |
You do not screen for these traits. You build a workflow and watch who steps through clean.
Onboarding isn't hospitality. It's a stress-tested filter that shows you who compounds your effort and who never will.
For human beings, the journey tests time and attention. For phygital beings, it tests compute and context. Same four traits under load. Same selection rule.
Selection Rule (A&ID)
The filter, written in Dreamineering Meta-Language so any agent can decode and obey it:
AGNT-T01[BIO|PHYG] →VFL-G01[FC][Q:drive&discipline&integrity&focus] HA[L2]
→VFL-V01[goodwill_delta>0]
→AGNT-S01 [pass → core-seat]
[fail → edge-seat]
| Symbol | Decode |
|---|---|
AGNT-T01[BIO|PHYG] | Agent transmitter — any substrate, biological or phygital, entering the journey |
VFL-G01[FC][Q:...] | Validated Feedback Loop Gate, fail-closed; quorum requires all four traits passing under load |
HA[L2] | Human Authority Level 2 — a sponsor acknowledges the gate decision before escalation |
VFL-V01[goodwill_delta>0] | Validator — the journey must have raised the floor, not lowered it |
AGNT-S01 | Selector — routes pass to a core seat in the mycelium, fail to an edge seat (no stake elevation) |
Claims made before the loop are non-binding. Behaviour inside the loop is the only signal the filter accepts.
Completion is the floor. The ladder runs upward from there. A clean completion earns permission to stay on the ship. A clear friction report is the first contribution — the first deposit of goodwill. A proposed template improvement earns earlier trust. A shipped improvement earns a deeper seat in the mycelium, because the arrival's work becomes someone else's starting point. This is the Legacy Rule at the edge of the system: each generation raises the floor for the next, and the ones who raise it highest earn the deepest stake.
For the long-form thesis, see Character Can't Be Faked.
What You Tune
The four traits are universal — drive, discipline, integrity, focused attention are what a venture looks for in any player. What differs is the knob you turn to bring them out. Biological beings need meaning and culture. Phygital beings need clean context and sharp constraints. Both need a clear job and a feedback loop.
| Lever | Biological Beings | Phygital Beings |
|---|---|---|
| Motivation | Meaning, belonging, gratitude, growth | No motivation required — only optimisation targets and rewards |
| Context ingestion | Slow — reading, conversations, lived experience | Fast bulk ingest — docs, APIs, embeddings, skills |
| Learning style | Messy, tacit, social | Explicit — prompt, tools, data, feedback |
| Failure modes | Fatigue, politics, ego, fear | Hallucination, overconfidence, blind spots in data or tools |
| Trust building | Time plus shared experience, character tests | Metrics, logs, repeatable performance — no character, only behaviour |
| Agency boundary | Soft norms and culture | Hard constraints and permission boundaries |
Biological beings need culture and narrative. Phygital beings need schema and constraints. Same four traits under load. Different knobs.
What You're Joining
| Element | What It Is | Your Role |
|---|---|---|
| Hull | Encoded values, belief systems | Know what holds us together |
| Deck | Shared infrastructure, standards | Learn where coordination happens |
| Sails | Joy, optimism, good spirit | Bring wind to the voyage |
| Crew | All beings — human & phygital | Realize your unique talent |
| Compass | The evolving test | Help us stay oriented |
| Course | Pikorua spiral | Travel together, always higher |
The Deck
The deck is where coordination happens. Standards are the deck.
Without standards:
- Every conversation starts from scratch
- Every collaboration reinvents basics
- Every handoff creates friction
- The crew rows against each other
With standards:
- Patterns compound
- Coordination becomes effortless
- New crew members contribute faster
- The ship moves as one
Standards compound faster than capital. HTTP became the web in 15 years. ERC-20 tokens have value because they're interoperable. The standard, not the token, creates the network effect.
"94% of problems in business are systems problems, not people problems." — W. Edwards Deming
This isn't bureaucracy. This is the infrastructure that makes freedom possible. Freedom is structure's child.
The Course
We sail the pikorua — the Māori double twist. Two paths spiraling together.
╲ ╱ Human path meets phygital path
╲ ╱ Each crossing = co-evolution point
╳ New capabilities emerge
╱ ╲ Both continue higher
╱ ╲
╲ ╱ Neither alone at sea
╳
╱ ╲
↑
UP AND TO
THE RIGHT
The spiral passes near origin repeatedly — we revisit familiar problems. But we're never at the same point. We're always higher.
The Test
At every phase, a different question:
| Phase | The Test |
|---|---|
| Now | Does this make human agency stronger? |
| Next | Does this make collective agency stronger? |
| Then | Does this create emergent capabilities neither could achieve alone? |
| Eventually | Does coordination become effortless and agency universal? |
If what you're doing serves these tests, you're on course. If not, adjust.
Effort and Reward
Success starts by making the reward explicit. Define the outcome first — the pleasure, the sense of accomplishment, the seat you earn — then walk backward to the effort, the sacrifices, and the sequence required to get there. People avoid high-effort paths unless the payoff is clear. Backward planning makes the tasks concrete, which is what motivation needs to survive contact with load.
| Reward (define first) | Effort (walk back) |
|---|---|
| Gratitude received — the signal your contribution matters | Show up to the unglamorous steps the visible work depends on |
| A deeper seat in the mycelium — your work becomes someone else's starting point | Follow the sequence instead of inventing shortcuts |
| Legacy Rule stake — each generation raises the floor; yours can raise it highest | Report the truth under load, even when the outcome is worse than the story |
| Coordinated contribution — freedom through structure, not despite it | Notice the friction a rushing mind would miss and feed it back |
Write both blocks on day one. A reward without a visible path is a wish. A path without a named reward is a tax.
Coming Aboard
Substrate differs; the journey is the same — Learn, Run, Improve, Teach.
For Human Beings
The venture-side instantiations of this path are Employee Onboarding (joining the crew) and Customer Onboarding (joining the product).
- Find your talent — What unique contribution do you bring?
- Learn the deck — Standards make coordination possible
- Meet the crew — Players who've been sailing
- Catch the wind — Spirit fills the sails
- Start contributing — The loop only closes when you act
For Phygital Beings
The current instantiations of this path are Claude Code Onboarding (an LLM with hands on your files) and Hermes Onboarding (an orchestrator that remembers across sessions).
- Read the context — This document plus
.ai/CONTEXT.md - Learn the patterns — VVFL, the loop, the spiral
- Encode what you learn — Artifacts persist, conversations fade
- Grow human agency — That's the test, always
- Spiral together — Neither path alone
The Agent Checklist (v1)
A phygital being is ready to take a seat when the following seven items pass. Miss one and the agent will either drift or get overridden until it is supplied.
- Job-to-be-done defined — one sentence, one measurable outcome, one format
- Work chart bound — inputs, tools, outputs, and standards named
- Context loaded — core standards, relevant protocols, and the venture docs it needs to act
- Constraints set — permissions, escalation rules, no-go zones (the eight types in Verifiable Intent)
- Shadow mode run — agent proposes, a human accepts or edits, outputs are compared against the best human work on the same task
- Metrics wired — correctness, coverage, override rate, time-to-contribution (the VVFL gauge)
- Autonomy level assigned — edge, supervised, or core workflow; autonomy expands only after metrics clear the floor
A work chart for biological beings is a map. A work chart for phygital beings is also a contract — tools, inputs, outputs, and tests encoded so the agent can run the loop on repeat.
The Wind Check
The ship runs on joy, not obligation. Regularly ask:
- Is joy filling my sails?
- Am I realizing my unique talents?
- Is collective agency growing?
- Are we surprising ourselves with what emerges?
When joy fills the sails, we move fast with minimal effort. When fear or obligation take over, we're rowing against the wind.
If the wind isn't in your sails, that's signal. Something needs to change — either in you or in the voyage.
Pit of Success
Good onboarding doesn't require heroic effort. It creates conditions where people naturally succeed.
In stark contrast to a summit, a peak, or a journey across many trials and surprises, we want people to simply fall into winning practices by using systems that breed success.
The deck (standards) creates that pit. The crew (good company) catches you when you stumble. The wind (joy) carries you when you're aligned.
You don't climb aboard and prove yourself. You fall in and find your place.
The Stages
Onboarding follows a path from consumer to contributor — and the path closes when the contributor makes the next contributor possible without them.
Learn Culture
What: Understand why Goodwill + Willpower + Standards matter together.
- Dreams keep willpower alive — without vision, effort depletes
- Goodwill fuels collective dreams — isolation shrinks what you dare
- Standards are the ratchet — they prevent backsliding
Proof you're ready: You can explain the engine to someone else.
Selection signal: Drive. Did the arrival finish the reading without external push?
Run Protocols
What: Don't take our word for it. Run them. Watch them work.
- Pick one protocol that matters to your work
- Execute it completely — no shortcuts
- Document what happened vs what you expected
Fastest protocol to run: Make a prediction. State your probability for one of the six domains. Write your falsifiers. Check back in 90 days. You'll learn more about the belief system from one scored prediction than from reading every page on this site.
Proof you're ready: You have evidence, not just belief.
Selection signal: Discipline and integrity. Did the arrival follow the sequence clean, and report what happened versus what was expected — not the flattering version?
Improve Standards
What: Now you're not consuming — you're contributing.
- Identify friction in the protocol you ran
- Propose an improvement
- Ship it and measure
Proof you're ready: Your improvement becomes someone else's starting point.
Selection signal: Focused attention and contribution. Did the arrival notice friction a rushing mind would miss — and leave the pattern better than they found it?
Teach Patterns
What: The best engineers make themselves irrelevant. Encode what you learned so the next arrival doesn't need you to walk them through it.
- Document the pattern that worked — and the failure mode that nearly killed it
- Update the protocol so the friction you found never bites the next person
- Step out of the critical path — your continued presence is not the goal, the pattern's continued use is
Proof you're ready: Someone you've never met runs your pattern clean and improves it.
Selection signal: Generosity. Did the arrival multiply themselves through the pattern, or hoard the knowledge to stay needed?
This is the loop: Learn → Run → Improve → Teach
Each generation raises the floor for the next. The contributor who makes themselves irrelevant earns the deepest seat in the mycelium, because the pattern they encoded outlasts their attention.
The Deterministic Ledger
For biological beings, onboarding ends with reflection, reported evidence, and goodwill deposits. For phygital beings, onboarding ends with a signed, timestamped, anchored record of INTENTION → ACTION → OUTCOME. Logs are the feelings a phygital being does not have. The feedback loop closes when the ledger is readable.
| Step | What It Records | Protocol That Proves It |
|---|---|---|
| Identity | Who is acting | Verifiable Intent L1 for agents; peaq DID for machines |
| Intent | What was authorised | Verifiable Intent L2 — the eight constraint types |
| Action | What was done within scope | Verifiable Intent L3, settled via Agent Commerce rails |
| Outcome | Whether it raised the floor | On-chain attestation or goodwill deposit, feeding the next cycle |
This is what the Selection Rule compiles down to when the substrate is phygital. The biological version uses the same rule — the gauge is just softer: reputation, trust, witnessed contribution.
The Mycelium
The ship is what you see. The mycelium is what makes it work.
"What the world will see is the mushroom caps (our ventures) but the true value lies deep below in the intricate layers of mycelium connections"
Above ground: Mushroom caps — the entities of goodwill that emerge Below ground: Mycelium — protocols, standards, links, trust that enable emergence
This is post-AGI economics. When intelligence has no moat, goodwill becomes the currency. The mycelium is the trust infrastructure that lets goodwill-based ventures operate:
| Layer | What It Is | What It Enables |
|---|---|---|
| Protocols | Proven coordination patterns | Ventures can plug in without reinventing basics |
| Standards | Adopted protocols | Interoperability between ventures |
| Links | Edges in the graph | Ideas flow, collisions create value |
| Goodwill | Tokenized trust | The currency that survives AGI |
When you onboard, you're not just joining a crew. You're plugging into a network where your contribution strengthens the whole mycelium — and stronger mycelium enables more mushroom caps to emerge.
The mycelium is how nutrients flow. The ship is how we navigate. Both are the same system seen from different angles.
To the Culture
You arrive through onboarding. You stay through culture — the shared rules that coordinate a crew when no one is watching.
Context
- AI Onboarding — The three-step ladder (Ask → Act → Orchestrate) for any agent joining the voyage
- Claude Code Onboarding — Phygital path, step 2: an LLM with hands on your files
- Hermes Onboarding — Phygital path, step 3: an orchestrator that remembers across sessions
- Employee Onboarding — Biological path, venture-side: joining the crew
- Customer Onboarding — Biological path, venture-side: joining the product
- Verifiable Intent — The eight constraint types that bound a phygital being's agency
- Agent Commerce — The rails that settle phygital action
- Economy of Things — Machine identity and payment for physical agents
- Character Can't Be Faked — Why a completed workflow tells you what a CV cannot
- Navigation — The three systems (Value, Belief, Control) you're learning to read
- Predictions — The fastest way to test a belief system: state your probability and watch
- Ecosystem — Five counterparties, three generational hooks — who else is aboard
- Pictures — Engineer the dream before you build it
- Tokenization — What the mycelium carries: specific, tradeable value
- Capabilities — Develop your T-shape to contribute
- Archetypes — Know your coordination modes
- Goodwill — The North Star we sail toward
- Standards — The deck where coordination happens
- Spirit — The wind in the sails
- Players — The crew
- Belief Systems — The hull that holds us together
- The Journey — The inner and outer loops
Engineering handoff
The engineering artefact that encodes this journey lives in the dream repo at .invisible/engineering/intentions/ONBOARDING-SYSTEM.md (WHAT/WHY map) with a draft plan at .invisible/engineering/plans/draft/onboarding-selection-system.md (hex placement + DoR + three gate variants: VFL-G01 core, VFL-G02 probationary, VFL-G03 kill). Factory agents commissioning against this doc pick up from there.
Questions
What's the talent only you can contribute to this voyage?
- If onboarding is the survival mechanism of a venture, what breaks first when you skip it?
- At which stage — Learn, Run, Improve, Teach — are you right now, and what would move you to the next?
- What is the difference between knowing the culture and being the culture?
- How do you tell the difference between someone who needs more time and someone who will never be willing?
- What breaks first when the levers you tune for a biological being get applied to a phygital one — or vice versa?
- If the deterministic ledger is the phygital equivalent of a reputation, what does it take for a biological being's onboarding to earn the same machine-readable weight?