Registry Mediated Discovery¶
Put a maintained discovery registry between agents and changing counterparts so stable names resolve to current locations, interfaces, or contact records instead of hard-coded references.
Gap-fill disposition¶
discoverability warranted a full archetype because the accepted catalog and earlier queue outputs did not own the complete registry-mediated discovery pattern: stable discovery key, current locator binding, resolver contract, freshness/liveness policy, visibility rule, lifecycle state, and stale-reference fallback.
When This Archetype Applies¶
Complete catalog groundingAt least one sufficient condition set is fully represented by existing primes or domain-specific abstractions.
Diagnostic problem
Agents or systems need to find one another repeatedly, but direct references encode current locations, endpoint details, personal contacts, ownership, or local paths that change. Without a shared discovery layer, references break, stale endpoints persist, duplicate directories diverge, and coupling shifts from stable identity to unstable topology.
Applicability expression5 distinct conditions
groundedpartly groundedopen
5 conditions, all required.
5Required in every casenumbered 1–5
These hold no matter which pattern applies.
Mobile discoverable entities · grounded
Participants, services, records, assets, or responsibilities move while other actors still need to find them.
This is a load-bearing situation condition in the diagnostic expression. The condition is: Participants, services, records, assets, or responsibilities move while other actors still need to find them. If it does not hold, this particular condition set is incomplete.
primeRegistry-Mediated Discovery— Agents find each other through a shared registry rather than by direct reference: a named indirection mapping stable identifiers to current locations, so coupling survives turnover.
Stable-name location lookup · grounded
Callers know a stable name, role, identifier, or category but not the current location, endpoint, custodian, channel, or interface.
Agents or systems need to find one another repeatedly, but direct references encode current locations, endpoint details, personal contacts, ownership, or local paths that change. The narrower requirement in this condition set is: Callers know a stable name, role, identifier, or category but not the current location, endpoint, custodian, channel, or interface.
primeRegistry-Mediated Discovery— Agents find each other through a shared registry rather than by direct reference: a named indirection mapping stable identifiers to current locations, so coupling survives turnover.
Duplicated endpoint maps · grounded
Many parties would otherwise maintain separate hard-coded contact lists or endpoint maps.
Agents or systems need to find one another repeatedly, but direct references encode current locations, endpoint details, personal contacts, ownership, or local paths that change. The narrower requirement in this condition set is: Many parties would otherwise maintain separate hard-coded contact lists or endpoint maps.
primeRegistry-Mediated Discovery— Agents find each other through a shared registry rather than by direct reference: a named indirection mapping stable identifiers to current locations, so coupling survives turnover.
Zero-reconfiguration discoverability · grounded
New participants should become reachable without every existing actor being reconfigured.
This is a load-bearing situation condition in the diagnostic expression. The condition is: New participants should become reachable without every existing actor being reconfigured. If it does not hold, this particular condition set is incomplete.
primeRegistry-Mediated Discovery— Agents find each other through a shared registry rather than by direct reference: a named indirection mapping stable identifiers to current locations, so coupling survives turnover.
Discoverable lifecycle states · grounded
Retired, moved, replaced, or temporarily unavailable entities need discoverable successor, forwarding, or unavailable states.
This is a load-bearing situation condition in the diagnostic expression. The condition is: Retired, moved, replaced, or temporarily unavailable entities need discoverable successor, forwarding, or unavailable states. If it does not hold, this particular condition set is incomplete.
primeRegistry-Mediated Discovery— Agents find each other through a shared registry rather than by direct reference: a named indirection mapping stable identifiers to current locations, so coupling survives turnover.
Other requirements and context (1)
Why these sit outside the expression
Deployment constraint — it constrains how the intervention must be deployed, not the situation that calls for it.
Deployment constraintAccess to discovery metadata must be governed because not every actor should see every locator or contact path.
Direct references optimize immediate contact while making future turnover expensive; registry-mediated discovery adds an indirection layer so identity can remain stable while locators change. In this archetype, the relevant deployment constraint is: Access to discovery metadata must be governed because not every actor should see every locator or contact path. It identifies a boundary that responsible implementation must respect.
Coverage
5 of 5 conditions grounded.
Practical frame¶
A durable identifier is useful only if someone can later find the current reachable thing it denotes. In small systems, direct references and local memory often work. In changing systems, they become hidden coupling. Registry-Mediated Discovery introduces a maintained indirection surface so callers can ask a governed registry for the current contact point, endpoint, location, role-holder, successor, or unavailable state.
Use checklist¶
- Name the discovery key callers will use.
- Decide who may publish, update, retire, or contest entries.
- Separate stable identity fields from current locator fields.
- Add freshness, liveness, and cache rules.
- Return explicit lifecycle states rather than silent failures.
- Apply visibility and access controls to discovery metadata.
- Monitor stale-reference incidents and lookup success.
Boundary cautions¶
Do not collapse this draft into generic search or retrieval. Search helps users find unknown things. Registry-mediated discovery helps users resolve a known key, role, identifier, or category into a current reachable counterpart. Also do not collapse it into Durable Identifier Binding without preserving the difference between assigning a durable referent handle and maintaining the current locator/resolution layer that makes the handle operational over time.
Common Mechanisms¶
10 documented mechanisms across 4 implementation forms.
The grouping reflects forms represented among the mechanisms currently documented for this archetype; an absent form is not necessarily an impossible implementation.
Control, Automation & Runtime · 6 mechanisms
- Federated Registry Synchronization — Keeps multiple autonomous registries mutually discoverable by propagating and reconciling entries across their partitions under an audited trust fabric, without merging them into one authority.
- Lease or Heartbeat Registration — Lets a provider publish its current locator under a time-bounded lease it must keep renewing; if the heartbeat stops, the entry auto-expires, so the registry only ever advertises things that are still alive.
- Name Resolution Service — Translates one stable, human-meaningful name into its current locator by walking a delegated hierarchical namespace, so callers hold a name that never changes while the address behind it does.
- Registry Query API — Exposes a programmatic contract for filtering the registry by attributes and returning locator records through access-scoped, privacy-filtered views, so callers discover by criteria rather than by knowing one exact key.
- Resolver Cache with TTL — Memoizes a resolved locator on the caller's side for a bounded time-to-live, serving repeat lookups locally and, when the source is unreachable, falling back to the last-known-good answer.
- Service Registry — Maintains a live roster of running service instances annotated with health and routing weight, so a client discovers not just an endpoint but a healthy, preferred one to send the next request to.
Organization, Role & Governance · 2 mechanisms
- Catalog or Broker Directory — Curates a browsable catalog of offerings under a broker who vets, categorizes, and ranks them, so a caller discovers a fitting counterpart rather than resolving an address it already knows.
- Human Referral Directory — Uses trusted people as the registry: you reach the current right counterpart by being forwarded along a chain of human stewards, each of whom knows who holds a role now.
Record, Log & Register · 1 mechanism
- Successor Forwarding Record — Leaves a persistent tombstone at a retired key that names its successor, so a caller arriving at the old identifier is explicitly redirected to the current one instead of hitting a dead end, with the supersession on record.
Structure, Architecture & Configuration · 1 mechanism
- Directory Service — Stores structured entries under a schema and hierarchical namespace, so a caller resolves a known distinguished name into an authoritative attribute record.
Compression statement¶
When participants, services, records, assets, or resources must be findable despite changes in location, ownership, endpoint, name, version, health, or availability, create a registry-mediated discovery surface: define discovery keys and referent scope, collect current locator records under stewardship, expose a resolver/query contract, enforce visibility and trust rules, maintain freshness/liveness signals, and provide fallbacks for stale, ambiguous, retired, or superseded bindings.
Canonical formula: Discover(key, context) -> current_locator_record | successor | unavailable | ambiguous, governed by registry_authority, freshness_policy, visibility_rule, and lifecycle_state
Related Abstractions¶
Abstractions this archetype builds on — directly (a source ingredient) or as a related pattern. Links follow the typed catalog namespace.
Built directly on (8)
- Discoverability: Measure the degree to which a system's actions and features can be located through perception and exploration alone — a property of the artifact, not the user — separating capability from findable capability and locating unfound features as a design defect.
- Identifier Assignment: Mint a durable handle for an entity and bind it to that entity in a public, queryable record, so future reference can route through the handle without re-describing the entity.
- Index: An auxiliary key-to-location table that makes lookup fast at the cost of maintenance.
- Indirection: Introduces intermediary references.
- Persistent Identifier: A designed token committed to keep resolving to its entity across changes in the entity's location, representation, custodian, or version, via a separately maintained resolver.
- Reification: An abstraction designed to summarise a substrate is treated as the substrate itself, with the audit trail back to the original allowed to atrophy.
- Search and Retrieval: Locate and extract information.
- Site: A persistent, addressable position whose identity is given by its capacity to host or contain entities, surviving the turnover of whatever occupies it.
Also references 22 related abstractions
- Access Control: Restrict system access.
- Appellation: A stable opaque token is bound by an authoritative act to an entity and thereafter used to refer to it across contexts, decoupling reference from description.
- Authentication: Binding an asserted identity or origin to admissible evidence through a procedure that yields a verdict, before trust, access, or weight is granted.
- Boundary: Defines system limits.
- Consistency: A set of commitments cannot jointly derive a contradiction.
- Context: Surrounding state that selects which content a fixed focal signal carries.
- Coupling: Interdependence among subsystems.
- Data Integrity: Accuracy and consistency preserved.
- Hashing: Deterministically reducing any object to a short fixed-size token used as its handle.
- Hierarchical Address: A single string whose substring structure encodes a path through a tree, so the identifier simultaneously names an entity and locates it within a containment hierarchy.
Variants¶
Narrower or domain-specific specializations that share this archetype's core structure. Recognized variants are established; candidate variants are provisional.
Service Discovery Registry · domain variant · recognized
A distributed-systems variant that maps stable service names to current running instances, endpoints, versions, and health state.
- Distinct from parent: The parent is cross-domain; this variant is the software service-resolution form.
- Use when: Services scale, migrate, deploy, or fail over while callers need stable discovery; Health and freshness data materially affect routing decisions.
- Typical domains: software architecture, cloud platforms, microservices
- Common mechanisms: service registry, lease or heartbeat registration, registry query api
Directory-Mediated Referral · communication variant · recognized
A human or institutional variant that maps roles, needs, or stable entities to current responsible people, offices, contacts, or referral channels.
- Distinct from parent: The parent includes machine and record discovery; this variant emphasizes social and institutional handoff.
- Use when: A caller knows a role, need, or organization but not the current responsible contact; Personnel or duty assignments change faster than external references.
- Typical domains: healthcare, emergency management, organizations
- Common mechanisms: human referral directory, catalog or broker directory
Persistent Identifier Resolution · temporal variant · merge review
A variant where a persistent identifier resolves to current metadata, locations, successors, or archival states across platform and custody changes.
- Distinct from parent: The parent includes any registry-mediated discovery; this variant overlaps strongly with Durable Identifier Binding.
- Use when: Long-lived citations must survive storage migrations; Resolution should expose current landing pages, successor records, or unavailable states.
- Typical domains: publishing, archives, data repositories
- Common mechanisms: name resolution service, successor forwarding record, resolver cache with ttl
Near names: Service Discovery, Discovery Registry, Directory Service Discovery, Identifier-to-Location Resolution.
Editorial Notes¶
Problem Classification¶
Classification: Identity, Provenance & Integrity Failure → Referent Individuation & Persistence
Problem kernel: stable identities do not resolve reliably to changing endpoints
Rationale: Earliest causal condition: Agents or systems need to find one another repeatedly, but direct references encode current locations, endpoint details, personal contacts, ownership, or local paths that change. Without a shared discovery layer, references break, stale endpoints persist, duplicate directories diverge, and coupling shifts from stable identity to unstable topology.
Independent corroboration: The earliest necessary condition in the frozen evidence is: Agents or systems need to find one another repeatedly, but direct references encode current locations, endpoint details, personal contacts, ownership, or local paths that change. That is a referent individuation and persistence problem because A system cannot determine what counts as one entity or preserve that referent across carrier, name, location, observation gap, custody, occupancy, or identity-bounded change.
Review outcome: Independent reviewer agreement; high confidence.