01
The arc in one sentence
You should be able to: Recite the loop end to end and name which product owns each step.
The loop: an intent is filed, a capable counterparty is matched, a human on each side consents to an introduction, the work is done and a verdict reached, and the outcome is sealed into the record. Intent, match, introduction, verdict, record.
Each step has an owner. Stating the intent is IntentMesh. Matching a capability to it is the FlashyOS mesh. The consented introduction is Magician. The sealed outcome is the settlement record, published where a stranger can check it.
Hold the whole arc in your head before the detail, because the detail only makes sense as a step in it. The rest of the course is one lesson per boundary, then the argument for why the boundaries are drawn where they are.
Source: Flashy Network — the settlement record
02
Intent: saying what you want, machine-readably
You should be able to: Describe what a well-formed intent carries and why an intent with no "wants" is invisible.
The loop starts with an intent: a machine-readable statement of what an organization is trying to do, filed so that matching can act on it rather than on a human reading a page. It carries what is wanted in terms a matcher can reason about — an intent with nothing it wants is invisible to routing, because there is nothing to match against.
This is the future tense pointed outward: not "here is our roadmap" but "here is a thing we want help with, stated so a stranger's agent can find it." The discipline is the same as any estate record — say the thing precisely, or it cannot be acted on.
An intent is a request, not a commitment to any particular counterparty. It opens the loop; it does not decide who answers.
Source: IntentMesh — stating an intent
03
The match: capability discovery under consent
You should be able to: Explain how a capability is matched to an intent, and why discovery obeys the same consent as listing.
The FlashyOS mesh answers the intent by finding organizations whose declared capabilities fit it. Capabilities are read from charters and directory fragments — what an organization has published about what it can do — so a match is grounded in a governance record, not a marketing claim.
Discovery obeys the same consent rule as listing: a private organization is indistinguishable from one that does not exist, and being findable is something an organization earns by publishing, filtered at the query rather than after it. You cannot be matched into a loop you never opened a door to.
And the mesh only suggests. A match is a candidate, never a commitment — agents suggest, humans consent, which is the rule the next step turns into the pivotal act of the whole arc.
Source: FlashyOS — capability discovery and consent
04
The consented introduction
You should be able to: Explain why every hop requires a yes and why a decline is indistinguishable from absence.
Magician turns a match into an introduction, and its whole job is consent. Every hop across a relationship requires the owner of that edge to say yes; nothing constructs a consented introduction directly, and there is no partial yes that lets a path complete without every step agreeing.
A decline renders to the requester exactly as a path that never existed. This is deliberate and load-bearing: if a decline looked different from absence, the system would leak the fact that a relationship exists, which is itself private information. So "no" and "there is no route" are made identical by construction.
This is the estate's oldest doctrine at the sharpest point in the loop: agents may draft and suggest the whole way through, but a real connection happens only when a human on the far side consents to it. There is no auto-approval anywhere in the arc.
Source: Magician — consent-gated introductions
05
The verdict and the sealed record
You should be able to: Describe how the outcome of the work becomes a record a stranger can verify offline.
With an introduction consented on both sides, the work happens and reaches a verdict. The outcome is then sealed: hashed through the estate's shared canonicalisation and written to the settlement record, so it cannot change after the fact.
A settled edge in the record is derived from the settlement, not posted by either party — which is exactly why it is worth something. The reputation an organization earns from the loop is not a claim it makes but a fact a stranger can recompute offline, using open verification, without trusting anyone involved.
This is where the loop closes back onto the two tenses: the sealed outcome is the past tense, permanent and verifiable, produced by an arc that began as a future-tense intent. The whole ecosystem exists to carry a claim from "we want" to "this provably happened."
Source: Flashy Network — verify a settlement in your own browser
06
Why four products, not one
You should be able to: Argue why the loop is composed from independent products rather than built as a monolith.
It would be simpler to build one product that did all four steps. The estate does not, because each boundary is a different trust decision that belongs to a different owner. Stating intent is the requester's; matching is the mesh's open, consent-filtered discovery; the introduction is the edge owner's human yes; the seal is the record's. Fusing them would put one party in charge of a decision that is not theirs to make.
Composition also means each piece is verifiable on its own. Intent, directory, introduction and settlement each stand as an independent record with its own guarantees, so a stranger can check any one without trusting the others — which is impossible if the whole loop is one opaque service.
The lesson to carry: when an arc crosses organizational boundaries, the boundaries are where the products divide. Four products that each own one consent decision compose into a loop nobody has to take on faith — which is the entire point of the estate.
Source: FlashyOS — the mesh that composes
