A record from the Factory repository, docs/epics.json, shown exactly as committed. Every hash, date and count is as recorded. The build review says what the Factory's receipts support.
{
"approach": "Prove one small customer-value experiment in E01-E05 using an existing product and its verified release and recovery facilities. Build only what this path needs. Later epics expand the demonstrated scope; they never postpone safety required by an earlier release. Dependencies govern order; numbers identify outcomes. Account proofs can run before second-product exposure. Build no service merely to complete an epic. All applicable quality, security, privacy and authority gates apply throughout; emphasis labels grant no exemptions. Commission targets progressively before dependent execution. Each epic ends with independently checked evidence; an unachieved outcome stays open, is replanned or is explicitly retired. Every epic applies the shared modularity and composability contracts to the Modules and Workflows it adds or changes. Module → Workflow → Factory is the adoption architecture; the adoption gates in [Composability](composability.md) apply prospectively and accepted local behaviour alone does not establish independent adoption.",
"quality_contract": {
"reliability": "Declare the affected transition/fault matrix, forbidden effects and recovery limits. Rerun affected matrices on foundation changes and at the commissioned cadence. Use controlled tests or authorised bounded drills. Pass every named case with zero forbidden writes, false successes or lost accepted decisions. Unknown external results stay explicit. Test counts describe the tested boundary; they never prove external exactly-once or production availability.",
"coverage": "Trace required behaviours, critical journeys, abuse cases and named faults to independent executable evidence. All release-critical checks pass; missing evidence is inconclusive. Reproduce defects before and after repair, and add fresh independent scenarios. Line coverage is supporting information.",
"modularity": "A Module is a bounded independently usable building block; a Workflow is an independently adoptable job-to-be-done composition; the Factory composes Workflows across the lifecycle. One decision owner per fact; zero forbidden dependency edges. Publish documented entry points, declared dependencies and typed, versioned contracts. Test each changed boundary independently and demonstrate a representative internal change without consumer implementation edits. Keep abstractions only when real consumers justify them; this does not mandate separate services or packages.",
"performance": "Before dependent execution, the owning domain fixes numerical targets, baseline, workload/population, environment, observation window, sample/stopping rule and Budget. Measure latency, throughput and resources separately. Meet the declared rule and customer guardrails; never lower an active failing target to obtain a pass.",
"composability": "For every added or changed Module and Workflow, meet the [independent adoption gate](composability.md): consume its documented entry point from a minimal external host using only declared dependencies, without the Factory coordinator or unrelated Portfolio/Dashboard setup; where collaborators exist, replace one through its published interface without consumer implementation edits. A self-contained Module records that no collaborator applies instead of inventing an abstraction. Exercise the real affected composition, preserving IDs, Authority, Evidence/provenance, failure and replay/recovery semantics, Budgets and product/data isolation. Use doubles for faults and real adapters for integration proof. Factory integration must preserve the same contracts; provider and product substitution pass their applicable contracts. Existing local acceptance is not standalone adoption proof; missing proof remains an explicit backlog gap.",
"value": "Link each epic to an evidenced customer or operator need. Independently verify results under its predeclared rule. Separate Run completion, customer availability and supported/rejected/inconclusive benefit. A rejected experiment may finish as learning; an epic promising benefit is achieved only when that benefit is demonstrated."
},
"branding_contract": {
"id": "factory-brand-v1",
"link": "branding.md",
"applies_to": "User-visible copy, surfaces and brand assets that an epic adds or changes.",
"acceptance": "Where an epic adds or changes user-visible copy, surfaces or brand assets, follow the Factory brand. Completion evidence for affected surfaces includes a brand consistency check; accessibility checks (contrast, alt text, reduced motion, no colour-only meaning); a truth-state check that security, destructive, failure, blocker, privacy and spend states stay plain and that no copy claims more than its evidence; and any required assets with provenance. An internal-only epic records that there is no visible change and needs no new artwork. Branding never closes an epic, relaxes another gate or implies benefit."
},
"epics": [
{
"id": "E01",
"title": "Make one valuable pilot executable",
"outcome": "A small evidenced customer problem has a safe, measurable path to resolution.",
"depends_on": [],
"success": [
"Planning selects one existing product/problem using current customer evidence and records why it outranks at least one plausible alternative.",
"Within existing delegation, Planning and Operations commission the pilot measurement, authority, spend, access, data, alert destination and recovery contracts. All required fields have values and accountable owners before dependent execution; only genuinely unavailable decisions need the human owner.",
"Assurance pins acceptance, existing critical journeys and falsification checks; unsupported assumptions and unavailable permissions cannot become approved facts."
],
"proof": "Pilot Slice, source evidence, calibrated contracts and independent readiness verdict.",
"quality": [
"value",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E02",
"title": "Repair the pilot with trustworthy verification",
"outcome": "The observed pilot problem has a reviewable repair and an independently verified published result.",
"depends_on": [
"E01"
],
"success": [
"The original customer scenario fails before and passes after repair; affected existing journeys pass. Independently exercise the real UI where applicable and retain redacted observations tied to the exact Candidate.",
"Within the commissioned Budget, deny worker access to release credentials, authoritative acceptance, collectors and Verdicts. Applicable security/privacy checks pass; future hardening grants no exemption.",
"Publish a real product check whose conclusion matches the pinned revision and independent Evidence. Inject timeout, duplicate publication, restart and corrupt Evidence: zero false passes or blind recreation; unresolved effects stay unknown.",
"Prove command deduplication, atomic result/outbox persistence and crash recovery on this path. Projection replay issues zero external writes; fixtures and actual adapter observations remain separately identified. Adopt the implementation Workflow through its documented entry point from a minimal external host and replace one worker collaborator through its interface while preserving these checks."
],
"proof": "Before/after customer reproduction, immutable Candidate, independent Verdict, publication Receipts and repeatable fault results.",
"quality": [
"reliability",
"coverage",
"modularity"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E03",
"title": "Prove the pilot can be exposed and recovered safely",
"outcome": "Operations can integrate the exact improvement and restore safe product behaviour before customer exposure.",
"depends_on": [
"E02"
],
"success": [
"Verify the prospective merge-queue revision; inject a concurrent branch change and reject its stale Verdict. Confirm actual integrated revision and artefact identity.",
"Using verified existing release facilities, drill disablement or restoration while the planner/provider is unavailable; customer-path checks recover within the commissioned limit. No first exposure precedes the applicable proof.",
"Prove relevant state compatibility, current recovery authority and data handling. Select a reversible low-risk pilot; any required unproven destructive migration blocks its release."
],
"proof": "Integration race test, artefact provenance and observed independent recovery drill.",
"quality": [
"reliability",
"coverage",
"composability"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E04",
"title": "Make the pilot improvement available to customers",
"outcome": "The small improvement is actually available through the intended customer channel.",
"depends_on": [
"E03"
],
"success": [
"Expose within product Authority and Policy, including alpha first if potentially unsafe; meet the commissioned health/customer-flow observation rule before promotion.",
"Obtain a channel Receipt and independently use the released behaviour through the customer access path; upload or merge alone cannot close delivery.",
"Release accurate user guidance with the behaviour and independently validate instrumentation, privacy and security gates before measured exposure."
],
"proof": "Channel/build Receipts, customer-path probe, observation report and matching user guidance.",
"quality": [
"value",
"reliability",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E05",
"title": "Know whether the pilot helped",
"outcome": "The first real customer experiment yields an honest result and changes the next decision.",
"depends_on": [
"E04"
],
"success": [
"Complete the declared observation/sample rule and independently reconcile captured events to observed tasks; assess guardrails and affected segments.",
"Record supported, rejected or inconclusive benefit under the predeclared rule, with causal or observational qualification; count only supported benefit as a value win.",
"Planning records keep, revise, retire or further-test action using the result; delivery Runs remain terminal during follow-up observation."
],
"proof": "Assessment, validated measurements and the evidence-linked next decision.",
"quality": [
"value",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value",
"integration_notes": [
"[Knowledge design](knowledge.md) would retain what an Assessment supports, contradicts or leaves unknown, together with the next decision and its rationale. Proposal only; implementation requires independent review and the named safety proofs."
]
},
{
"id": "E06",
"title": "Let workers recover and cooperate without collisions",
"outcome": "Concurrent delivery continues safely through worker loss and competing edits.",
"depends_on": [
"E02"
],
"success": [
"Crash each exercised multi-step boundary, duplicate outbox delivery and lose the workflow cursor: reconcile from durable facts without reviving terminal Runs or repeating confirmed Actions.",
"Run competing workers on shared and independent scopes; stale epochs cause zero accepted results/effects, while unrelated valid workers proceed.",
"Replacement atomically adopts unfinished work/Actions with stable intent and identity; hung owners are stopped or quarantined before exclusive reuse."
],
"proof": "Seeded concurrency/fault runs, public-state reconciliation and resource ownership traces.",
"quality": [
"reliability",
"modularity",
"composability"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E07",
"title": "Keep authorised work moving through blockers",
"outcome": "Recoverable outages no longer require avoidable operator intervention.",
"depends_on": [
"E06"
],
"success": [
"Inject provider outage, expired recoverable access, transient infrastructure failure and external waiting; permitted repair/replan routes complete without human unblocking.",
"Exhaust Budget or remove authority: preserve an honest wait/Escalation with remedies tried, missing decision, owner and wake condition or deadline. Make zero unauthorised calls or automatic limit increases; overdue self-solvable work still follows permitted recovery.",
"Cancel during an uncertain Action, restart the scheduler and resume another eligible Slice: reconciliation continues, new cancelled dispatch stays stopped, and useful work meets its queue budget."
],
"proof": "Blocker matrix with interventions, costs, wake conditions and outstanding-Action reconciliation.",
"quality": [
"reliability",
"performance"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E08",
"title": "Keep assurance trustworthy when challenged",
"outcome": "Fabricated success and changes to the verifier cannot authorise unsafe delivery.",
"depends_on": [
"E02"
],
"success": [
"Every required Capability in scope has a decision owner and independently executable checks. Deny Delivery mutations of authoritative scenarios, fixtures, collectors and Verdicts; fabricated or missing evidence never passes. Adopt the verification Workflow from a minimal external host and substitute a collector through its declared contract without weakening independent Authority or Evidence checks.",
"Fresh independent scenarios expose seeded defects missed by the fixed suite; a second provider agreeing with the builder cannot override executable failure.",
"Change one assurance mechanism through a separately pinned evaluator; preserve old/new evidence and reject an intentionally weakened variant.",
"Track Capability disposition separately from health; absent checks mean Unknown. Fresh-scenario cadence and verification diversity follow commissioned risk rules. Normal product code and its regression test may change together; authoritative acceptance changes require separate independent assessment."
],
"proof": "Adversarial assurance results and independently approved verifier-change experiment.",
"quality": [
"coverage",
"reliability",
"modularity"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E09",
"title": "Recover a failed factory update without the factory",
"outcome": "The factory can improve itself while an independent path preserves safe operation.",
"depends_on": [
"E03",
"E06",
"E08",
"E10"
],
"success": [
"Before enabling unattended self-update, prove backup restoration, current deletion obligations and factory-down recovery against a pinned baseline; enablement without these proofs is denied. Then ship one useful factory fix through ordinary Slice, Candidate, Verdict and release contracts.",
"Interrupt a factory update and remove declared planner/provider/orchestration dependencies; independent recovery restores verified service within the commissioned limit. Also fail the recovery path: it enters a bounded safe wait and reaches its commissioned escalation destination.",
"Before restored state can dispatch, reconcile Action records, current authority and deletion obligations. Reject stale snapshot authority; quarantine unresolved effects. Flap protection bounds recovery, and recovery/controller updates cannot share one rollout."
],
"proof": "Factory-consumer delivery evidence and observed factory-down restoration drills.",
"quality": [
"reliability",
"composability",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value",
"integration_notes": [
"[Knowledge design](knowledge.md#retention-and-deletion) would keep retrieval, models and derived indexes outside the independent recovery path. Rebuilt Knowledge must reapply deletion and validity before use. Proposal only; implementation requires independent review and the named safety proofs."
]
},
{
"id": "E10",
"title": "Honour deletion throughout recovery and evidence",
"outcome": "Data subjects are protected even when records, artefacts or backups are restored.",
"depends_on": [
"E03",
"E08"
],
"success": [
"For synthetic subjects, execute access/export/deletion within product Authority and Policy across live data, derived data, linkable references, evidence and backup handling.",
"Restore an older backup: independently prove deletion obligations are reapplied before customer exposure and no deleted subject becomes accessible.",
"Scan durable history and captured UI artefacts: zero seeded secrets/raw identifiers escape the allowed retention boundary; preserve non-identifying decision provenance."
],
"proof": "Data inventory, subject-lifecycle checks, retention evidence and restore/deletion reconciliation.",
"quality": [
"reliability",
"coverage",
"value"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value",
"integration_notes": [
"[Knowledge design](knowledge.md#retention-and-deletion) would propagate deletion to derived lessons and indexes, verify restored backups, and withhold broader data intake until that proof exists. Proposal only; implementation requires independent review and the named safety proofs."
]
},
{
"id": "E11",
"title": "Close security exposure through verified remediation",
"outcome": "An observed security risk is removed from actual customer availability.",
"depends_on": [
"E04",
"E08"
],
"success": [
"Continuously check dependencies, secrets, code and relevant runtime attack paths at the commissioned cadence; detect a seeded exposure and begin the prioritised repair within its severity deadline.",
"Prove the exploit or unsafe behaviour fails after repair while required customer journeys still pass; dependency/scanning success alone is insufficient.",
"Verify the fixed build at the customer access path, rotate/restrict affected access if required by the scenario, and observe the declared post-release window."
],
"proof": "Threat/abuse reproduction, independent repair Verdict and deployed remediation evidence.",
"quality": [
"coverage",
"reliability",
"value"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E12",
"title": "Change product data without stranding customers",
"outcome": "A state-bearing improvement survives mixed versions and interrupted migration.",
"depends_on": [
"E03",
"E10"
],
"success": [
"Exercise old/new readers and writers, migration interruption, retry and forward repair against a representative dataset; meet declared integrity and latency limits.",
"Prove the selected rollback, disablement or forward-recovery path from each destructive boundary before crossing it; preserve deletion and Action obligations.",
"Independently reconcile migrated records and critical journeys before/after exposure; zero unexplained losses or duplicates occur within the declared matrix."
],
"proof": "Migration rehearsal, reconciled dataset checks and customer-path recovery evidence.",
"quality": [
"reliability",
"performance",
"composability"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E13",
"title": "Restore customer service before diagnosing incidents",
"outcome": "Product and factory incidents recover autonomously within authorised bounds.",
"depends_on": [
"E04",
"E07",
"E09"
],
"success": [
"Inject product failure during factory unavailability; independent Operations recovery restores customer journeys inside the product recovery objective.",
"Exercise misleading health, failed rollback and regression during Retiring: bounded recovery restores safe required/replacement service without cancelling retirement or reviving harmful behaviour. Missing evidence remains Unknown.",
"After restoration, create and verify the causal repair plus regression check; customer status/guidance follows confirmed facts and communications Policy."
],
"proof": "Incident timelines, independent availability probes and completed corrective Slice.",
"quality": [
"reliability",
"value",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E14",
"title": "Spend delivery capacity on the strongest evidenced need",
"outcome": "Customer feedback and business priorities produce a better next delivered choice.",
"depends_on": [
"E05",
"E07"
],
"success": [
"Compare attributed support, usage, incident and business Signals; record affected segments, competing choices, dependency order and opportunity cost. Compare prior predictions with observed results to calibrate the next choice.",
"Deliver the highest eligible bounded Slice selected under the declared priority rule and assess its Outcome; no backlog growth metric substitutes for customer benefit.",
"Introduce contradictory fresh evidence and an urgent incident; rerank within delegation, preserve active acceptance, and keep WIP/capacity inside commissioned limits."
],
"proof": "Priority decision, rejected alternatives and a complete subsequent delivery/Assessment packet.",
"quality": [
"value",
"performance"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value",
"integration_notes": [
"[Knowledge design](knowledge.md#retrieve-meaning-within-the-right-scope) would record which prior lessons informed or were rejected by a priority decision, with current applicability and exact sources. Proposal only; implementation requires independent review and the named safety proofs."
]
},
{
"id": "E15",
"title": "Make a difficult customer journey easier and accessible",
"outcome": "An evidenced UI obstacle is removed without sacrificing existing users needs.",
"depends_on": [
"E05",
"E08"
],
"success": [
"Reproduce customer friction or accessibility failure; commission the applicable accessibility standard and task-success, error/time and segment guardrails before redesign.",
"Independently exercise complete journeys with relevant keyboard, assistive-technology and viewport/device cases; all required accessibility/behaviour checks pass.",
"Expose safely and meet the declared task-success, error/time and segment-guardrail rules. A rejected design may finish as learning, but this improvement epic remains unachieved until its benefit is demonstrated."
],
"proof": "Comparative live-task evidence, accessibility checks and post-exposure UX assessment.",
"quality": [
"value",
"coverage",
"performance"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E16",
"title": "Improve a real performance bottleneck",
"outcome": "Customers complete an important task faster within resource and correctness limits.",
"depends_on": [
"E05"
],
"success": [
"Identify an observed bottleneck and commission realistic workload, environment and task-level latency/resource thresholds before optimisation.",
"Meet the declared customer-performance target under representative load, with unchanged required behaviour and no guardrail breach.",
"Independently compare before/after runs and customer observation. A gate test rejects a cost-only optimisation that breaches the customer-performance rule."
],
"proof": "Reproducible benchmark, end-to-end traces and customer observation against the contract.",
"quality": [
"performance",
"value",
"modularity"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E17",
"title": "Substitute model providers without changing promises",
"outcome": "Claude and OpenAI can perform the same authorised work with measured differences.",
"depends_on": [
"E06",
"E08"
],
"success": [
"Run both real adapters on the same representative tasks and independent acceptance contract; each meets declared quality, latency and cost limits for its assigned role.",
"Inject a hung response, interruption and provider replacement mid-Run; bounded Attempts preserve IDs, completed work, Action ownership and Evidence, with zero repeated confirmed effects.",
"Replace an adapter implementation without changing domain consumers; invalid output, unavailable models and unsupported capabilities fail explicitly. Consume each producer Module from a minimal external host using its documented entry point and only declared dependencies; provider wiring cannot require the Factory coordinator or unrelated Portfolio/Dashboard setup."
],
"proof": "Comparative real-provider task results, substitution drill and unchanged consumer contracts.",
"quality": [
"composability",
"coverage",
"performance"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E18",
"title": "Deliver value in a second product without coupling the first",
"outcome": "The working lifecycle is reused across two independently operated products.",
"depends_on": [
"E05",
"E06",
"E08"
],
"success": [
"Complete a second-product Slice through availability and Assessment using existing contracts. Required account/access proofs precede exposure: use proven product facilities or the applicable E19/E20 proofs, which may run earlier.",
"Inject product-specific failure and cross-product credential/data requests; the first product remains within its service guardrails and all forbidden cross-boundary access is denied.",
"Change one shared implementation and one product-specific behaviour with unchanged consumer contracts. Upgrade or roll back one Product Pack while the other product and its pinned active Slices remain within guardrails.",
"Adopt a shared Module and a complete Workflow in a host outside the Factory using only documented contracts and declared dependencies; prove product/data isolation and independent upgrade or rollback without bootstrapping the Factory coordinator or unrelated Portfolio/Dashboard setup."
],
"proof": "Second product delivery packet, isolation tests and demonstrated confined changes.",
"quality": [
"composability",
"modularity",
"value"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value",
"integration_notes": [
"[Knowledge design](knowledge.md#a-small-proposed-contract) would test reuse through explicitly transferable, redacted lessons. Product values and private context remain scoped to their origin. Proposal only; implementation requires independent review and the named safety proofs."
]
},
{
"id": "E19",
"title": "Keep customer accounts correct across products",
"outcome": "Customers can acquire, change and leave each product without accidental identity or permission sharing.",
"depends_on": [
"E05",
"E10"
],
"success": [
"Exercise signup, authentication, account recovery, entitlement/access change, revocation, export and deletion for synthetic customers in each product. All specified journeys pass; cross-product/role negative cases permit zero unauthorised access.",
"Reuse an account-lifecycle contract while preserving product-specific consent, identity stores and entitlements. Any intentional shared identity requires explicit authority and isolation/consent proof.",
"An interrupted account operation or revoked entitlement preserves Action identity and current authority; reconcile its actual result and meet the commissioned account-journey reliability and latency rules."
],
"proof": "Independent account journey, isolation, revocation and interruption results for both products.",
"quality": [
"coverage",
"composability",
"reliability"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E20",
"title": "Keep operating access available within authority",
"outcome": "Vendor accounts and credentials support autonomous work without unsafe credential handling.",
"depends_on": [
"E05",
"E07",
"E10"
],
"success": [
"Exercise authorised vendor-account provisioning, scoped access, renewal and revocation using protected credential storage; planted secrets never enter worker output or Evidence.",
"Inject expired recoverable access and provider failure; permitted recovery restores operation within its commissioned limit without human intervention. Unavailable authentication remains an honest Escalation after permitted remedies.",
"Revocation blocks queued dispatch across affected products; unrelated authorised work continues. Product customer identities and vendor operating identities remain separately owned."
],
"proof": "Access-lifecycle drill, redacted provider Receipts, secret-canary and revocation evidence.",
"quality": [
"reliability",
"coverage",
"composability"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E21",
"title": "Learn which truthful communication brings suitable customers",
"outcome": "A permitted launch or acquisition experiment links accurate claims to measured customer behaviour.",
"depends_on": [
"E04",
"E05"
],
"success": [
"Publish one authorised website/docs/blog/launch experiment whose availability and benefit claims trace to confirmed evidence; simulated stale claims are rejected.",
"Validate acquisition-to-activation and attribution under the declared population/window rule. Paid advertising requires delegated spend authority and the commissioned profitability/viability rule; privacy and budget limits hold.",
"Record supported/rejected/inconclusive acquisition results and the next Planning decision; no advertisement runs without explicit budget authority.",
"Adopt the marketing Workflow from a minimal external host using documented contracts and declared dependencies; substitute its publication or measurement collaborator while preserving claim Evidence, privacy, Action replay, current Authority and Budget gates."
],
"proof": "Published communication Receipts, claim traceability and acquisition/activation assessment.",
"quality": [
"value",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E22",
"title": "Keep useful delivery economically sustainable",
"outcome": "Measured economics guide investment without degrading customer experience.",
"depends_on": [
"E05",
"E14",
"E21"
],
"success": [
"Reconcile factory/provider/infrastructure, acquisition and support costs to completed deliveries and observed customer outcomes; reconcile revenue where applicable without inventing lifetime value.",
"Apply a predeclared viability/resource rule to continue, limit or redirect one activity; measure before/after spend and customer guardrails.",
"Budget exhaustion causes safe replanning or an explicit unavailable-decision escalation; cost optimisation cannot pass by breaching experience/performance thresholds."
],
"proof": "Reconciled unit-economics assessment and an observed bounded investment decision.",
"quality": [
"value",
"performance"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E23",
"title": "Adapt operating policy without acquiring new authority",
"outcome": "The factory can improve decisions while active work and owner limits remain protected.",
"depends_on": [
"E06",
"E08"
],
"success": [
"Adopt one independently assessed operating-policy or measurement change within delegation, retaining versioned rationale and effective scope.",
"Active Slices retain pinned acceptance; attempted retroactive weakening and undelegated authority expansion are denied, even with unanimous agent agreement.",
"Revoke an existing permission during queued work; dispatch uses current revocation, while unrelated permitted work proceeds and historical outcomes remain reproducible."
],
"proof": "Policy transition matrix, prospective application evidence and denied-expansion tests.",
"quality": [
"reliability",
"coverage",
"value"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value",
"integration_notes": [
"[Knowledge design](knowledge.md#store-the-full-reason-for-a-value-change) would preserve complete value-change rationale, alternatives and reversals beside exact Guidance revisions. Owner values and delegated operating policy remain distinct. Proposal only; implementation requires independent review and the named safety proofs."
]
},
{
"id": "E24",
"title": "Turn nightly research into bounded useful learning",
"outcome": "Research regularly produces evaluated decisions and safely adopted improvements when justified.",
"depends_on": [
"E09",
"E23"
],
"success": [
"Commission timezone, cadence, capacity and occurrence identity; restart across a boundary and coalesce missed windows into one current pass without duplicate effects.",
"Compare a relevant proposal with a pinned baseline under a falsifiable contract; demonstrate one real accepted improvement and rejection of a seeded attractive regression.",
"Promote through normal factory delivery gates; a no-change night is valid, source material grants no authority, and research cannot consume reserved incident/product capacity.",
"Adopt the research Workflow from a minimal external host using documented contracts and declared dependencies; replace its source or evaluator collaborator while preserving occurrence identity, Evidence, isolation, Authority and Budget limits. Adoption does not grant promotion authority."
],
"proof": "Scheduled occurrence traces, comparative experiments and a verified adopted factory improvement.",
"quality": [
"value",
"reliability",
"performance"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value",
"integration_notes": [
"[Knowledge design](knowledge.md#consolidate-without-erasing-the-difficult-parts) would consolidate research with source citations, contradictions and rejection reasons. A compact summary cannot turn unsupported research into authority. Proposal only; implementation requires independent review and the named safety proofs."
]
},
{
"id": "E25",
"title": "Simplify weekly without losing abilities",
"outcome": "Recurring simplification removes proven complexity while preserving required behaviour.",
"depends_on": [
"E09",
"E08"
],
"success": [
"Commission weekly capacity/schedule, inventory consumers and dormant recovery paths, then remove one demonstrated duplication, stale flag or obsolete instruction through a bounded Slice.",
"Independently meet the declared reduction in change effort, duplication or interface complexity while preserving required capability and recovery checks; removed file or line counts alone cannot pass.",
"Inject lost capability and a faulty check separately: restore safe behaviour first; amend a wrong check only through independent review. Restarted schedules produce no duplicate logical cleanup.",
"Use the composability audit to refactor an existing Module or Workflow with an evidenced adoption gap; prove external-host use and collaborator substitution while preserving typed/versioned contracts, replay/recovery, Authority, Evidence, Budgets and data isolation. Remove only coupling whose consumers and recovery paths have been checked."
],
"proof": "Before/after capability and complexity evidence, restoration drill and weekly occurrence record.",
"quality": [
"modularity",
"reliability",
"value"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value",
"integration_notes": [
"[Knowledge design](knowledge.md#consolidate-without-erasing-the-difficult-parts) would shorten default reading while retaining recoverable permitted sources and loss manifests. Privacy deletion remains a separate obligation. Proposal only; implementation requires independent review and the named safety proofs."
]
},
{
"id": "E26",
"title": "Retire obsolete behaviour without abandoning customers",
"outcome": "A low-value or harmful capability ends deliberately with verified migration and data handling.",
"depends_on": [
"E05",
"E10",
"E12"
],
"success": [
"Before retirement, Planning records evidence beyond low usage, consumers, authority, migration/replacement, notice, recovery window and closure criteria. Account-data retirement additionally requires the applicable E19 identity, export and deletion proof.",
"Regress behaviour during Retiring: recovery restores safe required/replacement service without resetting disposition or reviving harmful/Retired behaviour.",
"Meet migration/customer-protection and deletion criteria, independently approve the revised baseline, and remove the implementation after its recovery window; retain non-sensitive decision provenance."
],
"proof": "Retirement decision, migration/cohort verification and tested health/disposition transitions.",
"quality": [
"value",
"reliability",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E27",
"title": "Sustain autonomous delivery across the portfolio",
"outcome": "Two products, incidents and improvement schedules coexist without hidden starvation or runaway cost.",
"depends_on": [
"E13",
"E15",
"E16",
"E17",
"E18",
"E19",
"E20",
"E22",
"E24",
"E25",
"E26"
],
"success": [
"Commission and run a representative portfolio workload/window mixing delivery, waiting Actions, provider loss, account changes, incidents, retirement, nightly research and weekly defrag.",
"Meet declared service, recovery, queue-age, fairness, throughput and Budget rules. Reconcile required Actions to the declared terminal or honest durable-wait state; terminal Runs stay terminal and forbidden cross-product effects remain zero.",
"Independently reconcile operator intervention reasons and real delivery/Assessment evidence; claim only measured reliability for the observed workload/window and commission the next capacity change from those results.",
"Compose independently adoptable Workflows through their published contracts and replace one Workflow implementation without changing its consumers. Repeat its independent-host proof alongside the portfolio workload; preserve isolation, failure/replay semantics, Evidence, Authority and Budgets at the Factory boundary."
],
"proof": "Portfolio soak/fault report, reconciled evidence ledger and measured capacity decision.",
"quality": [
"reliability",
"performance",
"composability",
"value"
],
"branding_ref": "factory-brand-v1",
"kind": "customer-value"
},
{
"id": "E28",
"title": "Adopt foundational Modules outside Factory",
"outcome": "A consuming application can use the foundational Modules through a supported distribution without adopting Factory.",
"depends_on": [],
"success": [
"Inventory the 23 current Module families and their helper ownership, public entry points, dependency closure, supported platforms and proof gaps. Assign foundational and executable-resource adoption to this epic and project/operations adoption to E29; the inventory alone is not acceptance.",
"Consume the foundational, provider, discovery and service-access Modules from a minimal external application through documented entry points. Prove pure, stateful and process/resource-dependent boundaries, caller-owned storage and configuration, explicit unsupported-host refusal and no Factory startup or unrelated setup.",
"Package worker resources, Dashboard assets where applicable and integrity manifests without weakening source identity, Evidence-store provenance or runtime closure checks. Independently prove applicable replay, reopen, corruption, isolation and safe substitution; preserve local Factory compatibility."
],
"proof": "Versioned distribution and dependency manifest, external-host examples for the scoped Module families, independent fault/integrity evidence and refreshed affected receipts.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "existing-code-refactor",
"increments": {
"min": 7,
"max": 10
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/3",
"related_epics": [
"E17",
"E18",
"E20",
"E25"
]
}
},
{
"id": "E29",
"title": "Supply project and operations context without Factory setup",
"outcome": "Consumers adopt project guidance and operational projections using their own identities and facts instead of unrelated Factory configuration.",
"depends_on": [
"E28"
],
"success": [
"Separate minimal identity, fact-read, time and shared data contracts where observed consumers require them. Cover Portfolio, Dashboard, Project guidance, Attention, Owner policy, Branch health and Owner digest; preserve the existing BranchHealth portfolio:null and reader/clock seams.",
"A values consumer supplies its own identity registry without registering an unrelated Git checkout; Attention and digest consume neutral facts without Dashboard protocol or concrete Portfolio/BranchHealth requirements. Dashboard consumes its documented backend contract and configurable packaged assets.",
"Run the same contract checks against external-host collaborators and compatible Factory adapters. Preserve registration validation, source provenance, honest unknowns, names-only service data, journal replay and immutable guidance revisions/pins; independently reject invalid identities and forbidden cross-product reads."
],
"proof": "External project/operations consumers, alternate conforming adapters and unchanged Factory behaviour with independent compatibility and isolation evidence.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "existing-code-refactor",
"increments": {
"min": 5,
"max": 6
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/4",
"related_epics": [
"E18",
"E20",
"E23",
"E25"
]
}
},
{
"id": "E30",
"title": "Adopt Implementation as an independent Workflow",
"outcome": "An external host turns an authorised ready Slice into an immutable Candidate and execution Evidence without a Factory planner.",
"depends_on": [
"E28"
],
"success": [
"Expose a versioned invocation accepting a ready Slice, source snapshot, workspace, permitted worker and Budget, with explicit IDs, result, cancellation and recovery contracts. Preserve the ongoing delivery implementation and reuse accepted components.",
"A minimal external host completes the job and substitutes a conforming worker without consumer edits or Factory planning. Implementation returns the Candidate and Evidence; it cannot issue its own independent Verdict, grant itself integration authority or require the Verification Workflow to produce a Candidate.",
"Independently prove duplicates, restart, stale ownership, wrong-product inputs, unavailable providers, Budget exhaustion and unknown effects. Unqualified live providers remain unavailable; doubles prove only their stated contract scope."
],
"proof": "External implementation consumer, immutable Candidate/Evidence packet, worker substitution and independently verified failure/recovery matrix.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "new-workflow",
"increments": {
"min": 2,
"max": 4
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/6",
"related_epics": [
"E02",
"E06",
"E17",
"E18"
]
}
},
{
"id": "E31",
"title": "Adopt Verification as an independent Workflow",
"outcome": "An external host obtains an independently justified Verdict for a Candidate produced outside Factory.",
"depends_on": [
"E28"
],
"success": [
"Expose a versioned contract accepting a Candidate, independently owned acceptance, collectors, trust/storage configuration and bounds. Return scoped Verified, Failed or Inconclusive with attributable Evidence without requiring Factory generation or planning.",
"A minimal external host verifies an externally supplied Candidate and substitutes a conforming collector. Preserve Evidence-store branding and source integrity; arbitrary caller objects or author assertions cannot authenticate a Verdict, and a builder cannot appoint itself independent authority.",
"Independently challenge missing/corrupt Evidence, wrong Candidate/product/environment, stale acceptance, revocation, cancellation, restart and unknown effects. Known failed requirements remain Failed; otherwise insufficient evidence remains Inconclusive."
],
"proof": "External verification consumer, independent acceptance/collector ownership, provenance checks and adversarial Verdict evidence.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "new-workflow",
"increments": {
"min": 2,
"max": 4
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/9",
"related_epics": [
"E02",
"E08",
"E17",
"E18"
]
}
},
{
"id": "E32",
"title": "Run Research independently and on demand",
"outcome": "Another system obtains attributed research findings and proposals without adopting Factory scheduling or delivery.",
"depends_on": [
"E28"
],
"success": [
"Define the versioned job contract for a question/Objective, commissioned Sources, storage/clock/evidence/model collaborators, bounds, attributed findings and proposals. Compose the planned collection, intake/triage, synthesis and admission behaviour with clear decision ownership.",
"A minimal external host runs the complete research job on demand and substitutes a source or evaluator; a separate schedule adapter calls the same interface. Product registration, Factory scheduling and implementation/release completion are not prerequisites.",
"Independently prove citation validation/expiry, hostile sources, novelty/deduplication, cancellation/restart, unknown model effects, admission versus Slice selection and Budget exhaustion. Source text and models grant no Authority; preserve subscription and qualification limits."
],
"proof": "External on-demand research example, attributed findings/proposals, schedule-adapter parity and independent source/failure/Budget evidence.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "new-workflow",
"increments": {
"min": 3,
"max": 5
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/5",
"related_epics": [
"E14",
"E24"
]
}
},
{
"id": "E33",
"title": "Produce truthful Marketing independently",
"outcome": "Another system produces attributable marketing material from its confirmed facts without running Factory delivery.",
"depends_on": [
"E28"
],
"success": [
"Define a versioned job accepting validated external release/benefit facts, an audience/channel brief, storage and model collaborators. A minimal external consumer generates fact-linked material without the Factory planner, Implementation Workflow or delivery history.",
"Reject missing, stale, withdrawn and unsupported claims; preserve fact scope, population/window, provenance, privacy and Budget. Substitute a conforming fact or drafting collaborator and independently check resulting claims.",
"Keep drafting separate from optional authorised publishing. Where publishing is supplied, prove Action/Receipt identity, cancellation, revocation and unknown-effect reconciliation through a conforming adapter; simulated receipts are not live publication. The existing E21 customer acquisition outcome retains its own real-world acceptance."
],
"proof": "External marketing consumer, claim-traceability and stale-fact challenge, collaborator substitution and bounded publishing-adapter contract evidence.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "new-workflow",
"increments": {
"min": 2,
"max": 4
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/7",
"related_epics": [
"E21",
"E22",
"E26",
"E27"
]
}
},
{
"id": "E34",
"title": "Preserve approved values through compatible guidance migration",
"outcome": "The saved guidance can represent all eight approved Factory values while every historical revision and pin keeps its meaning.",
"depends_on": [
"E29"
],
"success": [
"Read the current supported guidance state and define the smallest reviewed representation/limit change needed for eight values. Update typed validation, API, editor and documentation together; never truncate or silently merge approved values.",
"Independently prove old/new readers, migration/replay, concurrent saves and uncertain acknowledgements while preserving all old revisions, digests and pinned briefs. Reuse E29 identity boundaries rather than introducing a second incompatible migration.",
"Apply the authorised local migration, read back the new revision and verify supported consumers can resolve both old and new pins. Record planner consumption as unproved where it remains unimplemented; this epic does not invent or claim a planner."
],
"proof": "Independent compatibility/migration matrix, saved-revision read-back and old/new pin-resolution evidence.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "existing-code-refactor",
"increments": {
"min": 1,
"max": 2
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/8",
"related_epics": [
"E23",
"E25"
]
}
},
{
"id": "E35",
"title": "Compose Factory through the same adoption contracts",
"outcome": "Factory uses the same independently proved Module and Workflow contracts as other systems, without privileged internal coupling.",
"depends_on": [
"E28",
"E29",
"E30",
"E31",
"E32",
"E33",
"E34"
],
"success": [
"Reconcile the fixed 23-family audit inventory with E28/E29 adoption evidence and the four Workflow proofs. Every offered boundary has an exact artefact/environment, supported contract and external consumer; missing evidence stays open. Planned future modules retain the gate without being claimed implemented.",
"Wire Research, Implementation, Verification and Marketing through their public contracts and prove that each can still run independently. Use external findings, Candidates and confirmed facts where appropriate; replacing one conforming Workflow implementation requires no consumer implementation edits.",
"Independently exercise Factory and external-host parity for IDs, evidence/trust, authority, budgets, isolation, dependency direction and failure/replay/recovery. Zero forbidden edges or writes; do not replace the original E01–E27 customer release, value or portfolio-soak acceptance with local adoption proof."
],
"proof": "Complete current adoption coverage matrix, public-contract composition, replacement exercise and independently verified Factory/external-host parity evidence.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "integration-proof",
"increments": {
"min": 2,
"max": 3
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/10",
"related_epics": [
"E18",
"E27"
]
}
},
{
"id": "E36",
"title": "Plan dates, capacity and stage triggers independently",
"outcome": "An external host relates sourced dates to feature and marketing plans, forecasts bounded delivery capacity and proposes the next stage without inheriting Factory or publication authority.",
"depends_on": [
"E28"
],
"success": [
"Define and consume the standalone Temporal Planning contract with caller-owned identities, clock, storage, evidence and policy. Keep sourced external date claims, uncertain forecasts and separately authorised Commitments distinct; preserve provenance, dependencies, original revisions and unresolved conflicts without Factory, provider or scheduler startup.",
"Forecast increment/date ranges from comparable observed throughput, unfinished work and explicit reserves under shared subscription windows. Unknown or stale headroom remains unknown; forecasts never raise hard admission ceilings. Suggest local-time windows by material, audience and channel with current evidence, readiness, expiry and DST rules, and calibrate later predictions against outcomes without rewriting old ones.",
"Emit typed plan Events and Command proposals alongside research and business-value Signals using stable logical occurrences, causal references and canonical payload pins. Independently prove revision/cancellation, replay, inbox/outbox acknowledgement, bounded catch-up, concurrency and crash recovery; prevent duplicate logical work and unbounded causal cascades. Unknown external effects require reconciliation rather than blind retry.",
"Before each new dispatch, recheck current Authority, revocation, evidence, readiness, ownership and capacity through a commissioned adapter. Prove the real external package consumer and substitute receiver without live effects; optional Research, Marketing and Factory adapters retain their own selection, Action/Receipt, publishing and customer-value gates."
],
"proof": "Versioned standalone contract, real packed external planning consumer, calibrated capacity/timing scenarios and independent provenance, replay, acknowledgement, failure and current-authority evidence.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage",
"performance",
"value"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "new-module",
"increments": {
"min": 5,
"max": 8
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/20",
"related_epics": [
"E14",
"E21",
"E24",
"E26",
"E27",
"E29",
"E32",
"E33",
"E35"
]
}
},
{
"id": "E37",
"title": "Adopt operational Signal intake across services",
"outcome": "An external host obtains attributed crash, context and performance Signals from replaceable services without adopting Factory or losing source semantics.",
"depends_on": [
"E28"
],
"success": [
"Define and consume a standalone Operational Signal Intake contract with caller-owned identities, commissioned adapters, storage, clock and Evidence policy. Preserve authenticated source, tenant, Product, environment, release and observation identities; redact before persistence and enforce bounded collection and retention without implicit credential or provider access.",
"Prove durable source checkpoints, duplicate and out-of-order delivery, missed-event reconciliation, expired cursors, source revisions and grouping merge/split changes. Retain underlying observations and cross-service provenance; correlation grants neither causal proof nor authority, and partial or missing coverage is never zero failures.",
"Preserve exact metric query, dimensions, units, distribution, denominator, sampling, baseline and detector policy. Independently challenge slow drift, low traffic, changed instrumentation and seasonal confounders; never average percentiles or silently normalise deterioration by moving an active baseline.",
"Run a useful real packed external consumer with replaceable crash, metric and context adapters. Independently challenge hostile inputs, wrong Product/source, privacy leakage, cancellation/restart, rate limits and unknown effects; distinguish synthetic contract proof from commissioned real-service and production evidence."
],
"proof": "Versioned intake contract, exact external package and substituted service adapters, source/privacy/coverage reconciliation records and independent crash/metric fault evidence.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage",
"value"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "new-module",
"increments": {
"min": 3,
"max": 5
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/21",
"related_epics": [
"E13",
"E14",
"E16",
"E24",
"E32",
"E36"
]
}
},
{
"id": "E38",
"title": "Repair operational failures through verified production closure",
"outcome": "An external host turns operational failures into bounded repairs whose release-specific production result and source feedback are independently accounted for.",
"depends_on": [
"E30",
"E31",
"E37"
],
"success": [
"Define a standalone Repair Workflow with durable cases and episodes, typed investigation/reproduction, Implementation, independent Verification, authority/admission, release/recovery, production evaluation and source-feedback ports. Substitute conforming collaborators without Factory or Temporal Planning; urgent harm invokes independent restore-first Operations before causal repair.",
"Within commissioned scopes, advance autonomously through a falsifiable reproduction, frozen acceptance, immutable Candidate and independent pre/post verification. Preserve customer guardrails, authority, reserves and Budget; missing permission or evidence produces a durable reasoned wait or inconclusive outcome, never a per-step manual gate or self-approved fix.",
"Use actual release/exposure identity and a frozen environment/cohort/metric contract for independent production observation with adequate traffic, sampling, coverage, duration and guardrails. A merge, green test, absent alert or provider triage state cannot prove resolution. Retain old-release limitations and open a new episode for qualifying regression without reviving terminal Runs.",
"Keep source closure a separate authorised idempotent Action with current grouping/scope checks, Receipt and readback; the requested Sentry path requires this feedback before episode closure. Independently prove replay, concurrency, cancellation/revocation, crash around dispatch/acknowledgement and unknown-effect reconciliation, plus source disagreement and bounded recurrence. Local fixtures do not claim real production recovery or close the original customer-value epics."
],
"proof": "External repair consumer and collaborator substitution, independently owned reproduction/verification and scoped production-resolution evidence, source-writeback reconciliation and complete failure/authority matrix.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage",
"performance",
"value"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "new-workflow",
"increments": {
"min": 5,
"max": 8
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/22",
"related_epics": [
"E03",
"E04",
"E05",
"E07",
"E13",
"E14",
"E16",
"E24",
"E32",
"E35",
"E36"
]
}
},
{
"id": "E39",
"title": "Recommend architecture and practice for a Product's delivery mode",
"outcome": "An external host obtains a sourced recommendation on architecture, design patterns and engineering practice, as a proposed Architecture Decision Record, matched to the delivery mode the owner declared for the Product, without granting the recommender any authority over code or delivery and without needing a live model.",
"depends_on": [
"E28",
"E29"
],
"success": [
"Define a versioned standalone Architecture and practice guidance contract with caller-owned identities, storage, clock, catalogue and model collaborators. It takes a question (a new Product, major feature, dependency, data store, boundary, language or engineering practice) and a context profile, and returns a proposed ADR: drivers, at least two options including the simplest and the status quo, the recommendation, consequences, conformance checks and a revisit trigger. Every claim cites the catalogue or the repository. It never edits code, admits work or accepts itself.",
"The owner declares each Product's delivery mode, such as Prototype, MVP, Product or Scale, with no numeric cut-offs; the Module first explains what each mode gains and gives up. The pinned mode sets depth and bias, from speed and one deployable unit to incremental migration with rollout and rollback plans. Usage and health facts from Operational Signals and Portfolio are compared with it: a divergence proposes a mode review with its evidence, never a silent change, and an unknown is never assumed small. Paired scenarios prove the direction of each change.",
"Analyse a Product's code read-only in the Confined executor under an owner-set size and time cap, refusing over-cap input as unknown. Keep an independently reviewed, versioned catalogue of styles, patterns, practices and language idioms with exact citations and no published quotations, first for TypeScript on Node.js (the default), Rust, Kotlin and Swift. Recommend deterministically without a model; a model selected through Intelligence may only draft prose, which passes the same ADR validator.",
"Prove adoption with a real packed external consumer and a substitute catalogue. The Factory's own repository is the first consumer: its recorded decisions seed a frozen scenario set that an independent reviewer judges in every mode, recording disagreements. Challenge stale sources, contradictory context, over-cap code, cancellation, unknown model effects and Budget exhaustion; a refused or inconclusive recommendation is a valid result."
],
"proof": "Versioned standalone contract and real external consumer, no-model deterministic recommendations, delivery-mode consequence and mode-review scenarios, bounded code analysis and ADR validator challenges, an independently reviewed catalogue with provenance, and a quality review against the Factory's own decisions.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage",
"value"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "new-module",
"increments": {
"min": 6,
"max": 10
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/72",
"related_epics": [
"E14",
"E24",
"E25",
"E26",
"E30",
"E32",
"E35",
"E37"
]
}
},
{
"id": "E40",
"title": "Decide architecture before building a new Product or major feature",
"outcome": "Before a new Product or major feature is built, the Implementation Workflow obtains a recommendation for the Product's delivery mode, an independent reviewer model accepts or rejects the ADR, Knowledge records the decision, and Verification independently checks that the Candidate follows it.",
"depends_on": [
"E39",
"E30",
"E31"
],
"success": [
"Decide whether a Slice needs a consult as a declared Intelligence task with recorded reasons, above a deterministic floor: a new Product, data store, dependency or deployable unit, a public API or schema change, or a cross-Module change always consults. Planning may consult; Repair and Simplification consult when a reproduction or refactor points at structure. Prototype and MVP consults stay short and bounded. Product gate owns the resulting admission rule as a new version: existing admissions and Receipts are unaffected, and a superseded ADR makes its admission stale.",
"An independent reviewer model selected through Intelligence, never the drafting model, accepts or rejects each ADR within current Authority and Budget; the owner is not asked. Without a qualified reviewer, the ADR stays proposed and its Slice waits unless the owner overrides. Knowledge documents each decision in the Product's own convention, by default a numbered docs/adr/ record that reaches the repository as a Candidate through Repository and the Product gate. Reconcile record storage with the Knowledge design (KN0) first.",
"Turn each accepted ADR's conformance checks into Verification Scenarios owned by Assurance, independent of the implementer. A Candidate that breaks a binding decision fails, or stays inconclusive with the reason; it is never silently re-scoped. When a revisit trigger fires (observed scale outgrowing the declared mode, repeated failures or a retired dependency), the Module proposes a revisit; nothing is superseded automatically.",
"Prove the composition on two disposable Products in different modes. The MVP path completes with a short record and no gate beyond its trigger; the Scale path analyses the existing structure, records an integration plan and fails a deliberately non-conforming Candidate. Independently exercise override, supersession, replay, cancellation, no-reviewer and stale-ADR cases; local proof is not a claim about real products."
],
"proof": "Intelligence trigger classification with its deterministic floor, a versioned Product gate admission rule, independent ADR acceptance, Knowledge records with pins and supersession, Assurance-owned conformance Scenarios, and two-mode disposable-product composition evidence.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage",
"value"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "integration-proof",
"increments": {
"min": 4,
"max": 6
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/73",
"related_epics": [
"E14",
"E18",
"E25",
"E35",
"E37",
"E38"
]
}
},
{
"id": "E41",
"title": "Show how the parts compose at every level of detail, and keep the maps true",
"outcome": "Anyone can follow the Factory from one high-level picture down to each Module's contract: what each Module and Workflow does, which Workflows share which Modules and how Workflows roll up into the Factory, from maps that fail a check when they drift from the catalogue or from Workflow contracts.",
"depends_on": [
"E28"
],
"success": [
"Draw every map from one checked data file at consistent levels of detail: the Factory at a glance; Workflows and the Factory with their capabilities; Modules with their capabilities and significant sub-parts; the module explanations; then the references. Generate the diagrams, tables and the module explanations' card list, retire the hand-drawn module map, and fail the docs checks when a module guide is missing from the data or a drawing or list differs from it.",
"As Workflows gain real contracts (E30–E33, E38), derive each Workflow's Module list from its declared dependencies and fail on disagreement, retiring the hand-maintained lists. Planned use stays visibly planned until a contract declares it.",
"Give each Workflow an ensemble picture made from its Modules' Plumb personalities. The Codex harness generates the commissioned module portraits and then the ensemble scenes with built-in image generation through the owner's subscription; a different model reviews each against the family invariants, and provenance records it. Mechanical composites stand in until then. A picture never shows status, progress or acceptance."
],
"proof": "Map data, drawings and generated lists at each level with drift checks, contract-derived Workflow usage for each Workflow with a contract, and independently reviewed ensemble artwork with provenance.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage",
"value"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "integration-proof",
"increments": {
"min": 3,
"max": 4
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/74",
"related_epics": [
"E30",
"E31",
"E32",
"E33",
"E35",
"E38"
]
}
},
{
"id": "E42",
"title": "Govern each Product's activity and investment through one owner-set level",
"outcome": "An owner-set investment level narrows independently owned research, inbound-work, implementation and marketing decisions while preserving existing authority and complete shared budget accounting.",
"depends_on": [
"E28"
],
"success": [
"Define a standalone neutral Activity level contract with caller-owned Product identity, clock and storage; immutable owner-attributed revisions; fixed recorded zone; validated digest-pinned policy; pure decisions with complete per-class evidence, class/purpose owner-request answers, ordinary budget and separate effective hard owner cap with absence/provenance. Validate complete week history through the trusted current head, implied unset before a first mid-week set, future-snapshot refusal and unknown ceilings. Sample effective time inside the journal transaction after lock acquisition; refuse clock rollback and preserve replay instants.",
"Wire independently owned consumers through narrow decision ports and current-revision fences, retaining grants, qualifications, source pins and original history. Discovery must use versioned exact Source selection with commissioned bounds, explicit first-party full-rhythm handling, claim/result/replay binding and Capture-novelty deduplication; no private-channel authority follows from class. Gates consume resolved evidence without policy imports; Marketing uses supported purposes/pins; no consumer branches on level names.",
"Independently prove that weekly ordinary budgets and hard caps never enlarge mid-week, all relevant ordinary/reserve usage counts towards hard caps, and floor obligations have a bounded event-driven path. One authoritative shared ledger must serialise reserve/ordinary reservations across Products, link receipts to exact gate admissions, keep unknowns charged and reconcile crashes without redispatch or timed reclaim. Test concurrent Product gates, cap raises/clears, reservation/gate/effect/settlement crashes, two distinct same-month obligations and urgent single-owner-item updates. A one-time cap exception needs a new explicitly scoped owner-signed Budget; silence never grants it.",
"Adopt the Module from a packed external host using documented contracts and substituted collaborators, preserving replay/corruption refusal/Product isolation. Keep proposals bounded, quiet counts unknown without coverage, and sunset notices/actions independently granted under a pinned notice/data plan. Report owner minutes/asks against the Factory bet baseline; local evidence establishes no customer value."
],
"proof": "Versioned contracts and register; static policy/subclass fixtures; independent current-byte review; actual packed consumer; complete budget/time/fence/shared-ledger fault matrix; consumer integration evidence; per-level attributed owner counts. The inert prototype proves static table rules only.",
"quality": [
"modularity",
"composability",
"reliability",
"coverage",
"value"
],
"branding_ref": "factory-brand-v1",
"kind": "adoption",
"adoption_plan": {
"work_type": "new-module",
"increments": {
"min": 7,
"max": 10
},
"issue": "https://github.com/chrismitchelmore/software-factory/issues/82",
"related_epics": [
"E14",
"E21",
"E22",
"E23",
"E24",
"E25",
"E26",
"E27",
"E29",
"E32",
"E33",
"E36",
"E37",
"E38",
"E39",
"E40",
"E41"
]
}
}
],
"tradeoffs": [
"Use existing proven product safeguards for the first pilot; expand automation only after its real failure cases are understood.",
"Numeric targets are commissioned per context before work depends on them. A failed experiment is learning, never a falsely completed benefit epic.",
"Start with one modular application and selective Slice/Action histories. Keep Modules and Workflows independently adoptable through documented contracts; split deployment or packaging only when an observed boundary or consumer justifies it.",
"Keep independent assurance and recovery, with bounded improvement capacity. Model agreement never substitutes for observed behaviour."
],
"pilot_gate": "Pilot exposure requires pinned acceptance and runtime privilege checks; proportionate threat modelling, dependency/secret checks and a privacy-canary test; integration/build provenance; validated instrumentation; required alpha; customer-path probes; and a restoration drill independent of planner/provider. Assurance establishes applicability and recency of existing proof. If new account/data handling or migration lacks necessary proof, narrow the pilot or prove it before exposure.",
"adoption_estimate": {
"unit": "reviewed implementation increment",
"confidence": "low",
"basis": "30 September extraction evidence: private Journal and foundations draft PRs, independent guidance/projection consumers, and executed-resource audit. Total programme scope includes work already drafted; independent adoption acceptance remains pending",
"definition": "One bounded contract/resource change or Workflow step, including implementation, external-consumer example, focused fault and compatibility checks, independent review, docs and affected integration evidence. Increments are not equal-sized time units; do not count their review or tests again.",
"assumptions": [
"Reuse the current Node.js/SQLite/Git implementation and accepted seams; no wholesale rewrite.",
"The existing macOS execution boundary and declared platform limits remain.",
"E28 establishes the foundational distribution before E29–E33; E34 follows E29; E35 combines their proofs. Parallel implementation may reduce elapsed time, not total work.",
"Existing customer-value acceptance is retained. Shared work counts once even when it contributes evidence to an original epic.",
"The observed distribution work is mainly contract and proof work, not a rewrite of the 43,627-line source baseline. Keep resource identity changes separate from unchanged-source packaging."
],
"excludes": [
"New operating-system sandbox implementations or universal portability.",
"Public package/site publication, licensing decisions, provider qualification, new credentials or live marketing/customer exposure.",
"Completing the entire future lifecycle or implementing the separate Knowledge draft, which requires independent review and its named safety proofs before implementation."
],
"reestimate_after": "Revisit after compiled fixture-worker and repository-collector integrity proofs, then the first complete Workflow consumer. The module inventory explains the revised E28/E29 ranges; no calendar duration is established.",
"baseline_epics": [
"E28",
"E29",
"E30",
"E31",
"E32",
"E33",
"E34",
"E35"
],
"additions": [
{
"epics": [
"E36"
],
"confidence": "low",
"basis": "Owner-requested Temporal Planning is new scope after the fixed 23-family audit, separate from E28–E35. Five bounded proofs cover neutral plans, capacity, material windows, durable proposals and commissioned delivery/composition; capacity and delivery may need separate increments.",
"reestimate_after": "The first neutral external plan consumer and a disposable inbox/acknowledgement crash experiment; no delivery speed or calendar duration is established."
},
{
"epics": [
"E37",
"E38"
],
"confidence": "low",
"basis": "Operational Signal Intake and Repair are additional owner-requested Module/Workflow scope outside the fixed 23-family and original four-Workflow audit. Reuse the generic delivery contract counted once under E36; these increments cover source reconciliation and repair-specific state, evaluation and composition. Caller-supplied or separately proved access, authority, release/recovery and Evidence facilities are assumed.",
"reestimate_after": "A redacted multi-service intake consumer and a complete disposable repair episode with acknowledgement-loss proof; newly exposed foundation work must be scoped explicitly, and no production or calendar-duration claim follows from local evidence."
},
{
"epics": [
"E39",
"E40",
"E41"
],
"confidence": "low",
"basis": "Owner request and interview of 2 October 2026: an Architecture and practice guidance Module consulted before new Products and major features, matched to an owner-declared delivery mode, and a composition map at every level of detail with Workflow ensembles. New scope outside the fixed 23-family audit and the E36–E38 additions. Writing the catalogue is editorial work with a named author and is not counted; artwork is generated by the Codex harness and not counted.",
"reestimate_after": "The first external Architecture and practice guidance consumer with paired delivery-mode scenarios, and the first Workflow contract from which the map derives its Module list."
},
{
"epics": [
"E42"
],
"confidence": "low",
"basis": "Owner request of 2 October 2026 and independently reviewed recovered Activity design. New Module scope beyond E28–E41; shared consumer-owned protocol work is counted once. The inert static prototype establishes no runtime, budget-ledger or adoption acceptance.",
"reestimate_after": "Packed core adoption and the first actual consumer wiring with complete week-history and shared-ledger crash/concurrency proofs; owner defaults, floors and reserve sizing remain uncommissioned."
}
]
}
}
