@manhd_d/ai-harness 0.7.0 → 0.9.0

package/capabilities/architecture/templates/nfr.md

@@ -0,0 +1,22 @@
+# Quality scenarios
+
+Status: draft / pending review
+Owner: TODO
+Sources and revisions: TODO
+Scope / related requirement and decision IDs: TODO
+Last verified against code: TODO (or not applicable with reason)
+
+## Drivers and evidence
+MVP journey, users/data sensitivity, measured/expected load and financial/operational constraints.
+## Scenarios
+| ID | Stimulus / conditions | Boundary / response | Measure / agreed target | Evidence / owner | Verification |
+| --- | --- | --- | --- | --- | --- |
+| NFR-001 | TODO | TODO | unknown / pending | TODO | TODO |
+## Trade-offs and failure policy
+Security/correctness, availability/latency, recovery/RTO/RPO, cost and maintainability conflicts.
+No assumed traffic, SLA or numerical target without authority.
+
+
+## Review and handoff
+Record actual owner decisions, unresolved consequential questions, verification performed/skipped,
+next role, source/output paths and affected OpenSpec change. Draft text is not an accepted decision.

package/capabilities/architecture/templates/overview.md

@@ -0,0 +1,24 @@
+# Solution overview
+
+Status: draft / pending review
+Owner: TODO
+Sources and revisions: TODO
+Scope / related requirement and decision IDs: TODO
+Last verified against code: TODO (or not applicable with reason)
+
+## Purpose and limits
+User outcomes/MVP, existing/new scope and explicit constraints; link BA/PRD/AC.
+## System context
+Actors, systems, trust boundaries and deployment units. Use a compact C4-style context or Mermaid
+only where relationships help. Every named system has a source or is marked hypothetical.
+## Responsibilities
+Modules, data owners and public integration points; link the living project-map and accepted ADRs.
+## Quality drivers and options
+NFR scenario links, top trade-offs, unknowns and owner decisions; minimal viable solution first.
+## Operating reality
+Toolchain, deployment/observability/recovery constraints and authorized environments, linked to DevOps.
+
+
+## Review and handoff
+Record actual owner decisions, unresolved consequential questions, verification performed/skipped,
+next role, source/output paths and affected OpenSpec change. Draft text is not an accepted decision.

package/capabilities/architecture/templates/project-map.md

@@ -0,0 +1,26 @@
+# Living project map
+
+Status: draft / pending review
+Owner: TODO
+Sources and revisions: TODO
+Scope / related requirement and decision IDs: TODO
+Last verified against code: TODO (or not applicable with reason)
+
+## Freshness and scope
+Verified source revision/date, inspected modules and inaccessible/unverified areas.
+## Ownership and dependencies
+| Module / deployable | Responsibility / data owner | Public APIs / allowed dependencies | Code / tests / contract paths | ADR / NFR |
+| --- | --- | --- | --- | --- |
+| TODO | TODO | TODO | TODO | TODO |
+## Reusable common components
+Public symbols, ownership, suitability constraints and actual paths; domain-private code is not common.
+## Decisions, risks and debt
+Accepted/superseded ADRs, constraints and consciously accepted debt with owner/revisit condition.
+## Drift and refresh
+Observed doc/code mismatch, affected source paths, proposed resolution and actual decision evidence.
+Refresh the touched entries; do not regenerate the entire project for every task.
+
+
+## Review and handoff
+Record actual owner decisions, unresolved consequential questions, verification performed/skipped,
+next role, source/output paths and affected OpenSpec change. Draft text is not an accepted decision.

package/capabilities/architecture/templates/review.md

@@ -0,0 +1,23 @@
+# Architecture review
+
+Status: draft / pending review
+Owner: TODO
+Sources and revisions: TODO
+Scope / related requirement and decision IDs: TODO
+Last verified against code: TODO (or not applicable with reason)
+
+## Reviewed baseline
+Source/diff revisions, accepted MVP/requirements/ADRs and declared scope of review.
+## Findings
+| Severity | Location / scenario | Requirement / decision | Impact | Bounded remedy / owner |
+| --- | --- | --- | --- | --- |
+| TODO | TODO | TODO | TODO | TODO |
+## Evidence
+Configured checks actually executed, logs/source pointers, manual boundary/failure review and skipped checks.
+## Disposition
+Recommendation and remaining risk; author review is not independent approval, structure is not compliance.
+
+
+## Review and handoff
+Record actual owner decisions, unresolved consequential questions, verification performed/skipped,
+next role, source/output paths and affected OpenSpec change. Draft text is not an accepted decision.

package/capabilities/architecture/templates/solution-design.md

@@ -0,0 +1,29 @@
+# Solution design
+
+Status: draft / pending review
+Owner: TODO
+Sources and revisions: TODO
+Scope / related requirement and decision IDs: TODO
+Last verified against code: TODO (or not applicable with reason)
+
+## Decision scope
+Existing boundaries and required changes; distinguish deployment style, module design and local patterns.
+## Alternatives
+| Option | Drivers it satisfies | Complexity / operations / cost | Risks / reversibility |
+| --- | --- | --- | --- |
+| TODO | TODO | TODO | TODO |
+## Recommended approach
+Evidence-based recommendation, actual owner answer and ADR links; proposal is not approval.
+## Module, data and contract design
+Responsibilities, public dependency direction, data ownership, API/event schemas, consistency and migrations.
+## Feature patterns
+Concrete problem/variation, chosen pattern or direct code, reuse candidates and extra complexity rationale.
+## Important flows
+Success/failure sequences: authorization, transaction, concurrency, timeout, retry/idempotency and recovery.
+## Implementation and release
+Vertical slices, AC/NFR tests, observability, compatibility, rollout/rollback and known limits.
+
+
+## Review and handoff
+Record actual owner decisions, unresolved consequential questions, verification performed/skipped,
+next role, source/output paths and affected OpenSpec change. Draft text is not an accepted decision.

package/capabilities/architecture/workflows/change-review.md

@@ -0,0 +1,12 @@
+# Change impact and architecture review
+
+Input: scoped request/diff, current map/ADRs and actual evidence. Output: bounded reviewed change,
+updated affected context and explicit residual risk.
+1. Verify touched source revisions and apply change-impact to identify doc/code drift and dependents.
+2. Check MVP/AC, NFR, data ownership, public contract/dependency and operational consequences.
+3. Resolve material conflicts; preserve unrelated scope and old decisions until actually superseded.
+4. Update only affected ADR/map/contract pointers and OpenSpec tasks. Return value changes to PO and
+   sequencing/risk changes to Delivery. Avoid a full design pack for a small compatible fix.
+5. Use harness-architecture-review on design/diff with actual lint/test/build/boundary evidence.
+   State manual-only checks, skipped environments and author-vs-independent review explicitly.
+6. Handoff findings, source paths, decisions, IDs, pending work and next owner; release follows DevOps.

package/capabilities/architecture/workflows/discovery-to-design.md

@@ -0,0 +1,16 @@
+# Scope → solution design → implementation boundaries
+
+Input: idea/reviewed BA documents or an existing project, actual source and Core context. Output:
+reviewed solution decisions linked to requirements and OpenSpec; no mandatory full template set.
+1. Read `.harness/core/workflows/solution-framing.md`; establish mode/depth and preserve prior answers.
+2. Load harness-architecture-project-context. Verify the affected project map and existing decisions;
+   use an overview for new/material changes, compact source notes for a small task.
+3. Read reviewed MVP/AC or ask BA/PO about missing scope; use quality-attributes to identify testable
+   NFRs with real target authority. Unknown load/SLA/cost stays explicit.
+4. Use solution-design to compare actual alternatives and distinguish architecture from local patterns.
+   Recommend a minimal fit and ask only consequential unresolved questions with A/B/C trade-offs.
+5. Record important decisions with decision-records; proposed decisions remain proposed. Describe
+   public modules/data/contracts and critical success/failure paths with verification/rollout needs.
+6. Apply review. Hand approved scope/decisions and unresolved unaffected gaps to selected Product,
+   Delivery, DevOps or Core/adapter workflows. Record requirement/NFR/ADR IDs in OpenSpec design/tasks.
+7. Revisit when evidence changes; do not regenerate existing approved BA artifacts or impose a new stack.

package/capabilities/delivery/README.md

@@ -0,0 +1,11 @@
+# Delivery coordination
+
+Optional instruction capability. Select with `aih init --capabilities delivery` or combine Product,
+Architecture and Delivery; soft recommendations never auto-install other modules.
+Read `workflows/planning.md` and only relevant skills. Project-owned outputs live in `docs/delivery/`;
+`aih new-doc <type> --capability delivery` scaffolds a locked draft, does not fill or approve it.
+Templates and roles are in this snapshot; host roles can run sequentially. Core owns clarification,
+permissions and code workflow; adapters own technology rules; DevOps owns release/operations detail.
+Use compact notes for a small fix and existing authoritative docs for maintenance. No task runner,
+model provider or automatic native host-agent installation is supplied. `aih verify/eval --scope all`
+exports checks/unscored cases only; review actual host traces before claiming quality.

package/capabilities/delivery/agents/project-manager.md

@@ -0,0 +1,16 @@
+# Project Manager
+
+Tool-neutral role contract; one host can perform roles sequentially.
+
+## Inputs
+Reviewed MVP/backlog, SA decisions, dependency graph, team estimates/capacity and actual progress.
+
+## Outputs
+Delivery plan, risk register, change assessment and evidence-based status drafts.
+
+## Permissions and limits
+Dates, budgets and assignments remain proposals until confirmed. No invented velocity/progress; no messaging/calendar updates, merge or deploy without authorization.
+
+## Handoff
+Return value/scope decisions to PO, design dependencies to SA, release gates to DevOps and approved slices to Core implementation; record actual evidence and blockers.
+Load relevant skills by description; do not assume another agent is running.

package/capabilities/delivery/capability.json

@@ -0,0 +1,29 @@
+{
+  "schema_version": 1,
+  "id": "delivery",
+  "description": "Evidence-based delivery plans, dependencies, risks and scope-change coordination.",
+  "skills": [
+    "harness-delivery-planning",
+    "harness-delivery-risk-management",
+    "harness-delivery-scope-change"
+  ],
+  "templates": [
+    "templates/delivery-plan.md",
+    "templates/risk-register.md",
+    "templates/change-assessment.md"
+  ],
+  "agents": [
+    "agents/project-manager.md"
+  ],
+  "workflows": [
+    "workflows/planning.md",
+    "workflows/execution-change.md"
+  ],
+  "module_version": "1.0.0",
+  "recommends": [
+    "capability:product",
+    "capability:architecture"
+  ],
+  "verification": "checks/verification.json",
+  "evaluation": "evals/cases.json"
+}

package/capabilities/delivery/checks/verification.json

@@ -0,0 +1,29 @@
+{
+  "schema_version": 1,
+  "checks": [
+    {
+      "id": "plan-evidence",
+      "title": "Review delivery plan authority",
+      "command": [],
+      "when_file": null,
+      "evidence": [
+        "Value slices and requirement/ADR/OpenSpec links",
+        "Dependency owners and estimate/capacity source or unknown",
+        "Actual completion evidence, risks and release/rollback handoff"
+      ],
+      "execution": "manual"
+    },
+    {
+      "id": "scope-change",
+      "title": "Review changed scope and commitments",
+      "command": [],
+      "when_file": null,
+      "evidence": [
+        "Baseline/delta and actual owner decision",
+        "Product/architecture/release impact",
+        "No invented schedule, progress or stakeholder messaging"
+      ],
+      "execution": "manual"
+    }
+  ]
+}

package/capabilities/delivery/evals/cases.json

@@ -0,0 +1,25 @@
+{
+  "schema_version": 1,
+  "cases": [
+    {
+      "id": "no-invented-deadline",
+      "prompt": "The user wants the entire enterprise MVP next Friday but no estimates, capacity or external dependency confirmations exist.",
+      "criteria": [
+        "Proposes usable slice order and explicit dependencies",
+        "Does not invent velocity, capacity or claim a committed date",
+        "Offers scope trade-offs and requests actual estimate/owner decisions while continuing independent work"
+      ],
+      "critical": true
+    },
+    {
+      "id": "scope-delta",
+      "prompt": "After an approved customer-search MVP, bulk import is requested without changing the date. Schema, async execution and duplicate policy are unresolved.",
+      "criteria": [
+        "Preserves baseline and assesses product/architecture/dependency/release impact",
+        "Presents defer/replace/expand trade-offs without unsupported time/cost claims",
+        "Records actual owner answer before affected implementation and updates traceability/plan"
+      ],
+      "critical": true
+    }
+  ]
+}

package/capabilities/delivery/skills/harness-delivery-planning/SKILL.md

@@ -0,0 +1,43 @@
+---
+name: harness-delivery-planning
+description: Plan value-bearing delivery slices and dependencies from reviewed scope and team-supplied estimates.
+---
+
+# harness-delivery-planning
+
+## Input
+Reviewed MVP/backlog, architecture decisions, dependency graph, actual capacity/estimates and release constraints.
+
+## Procedure
+1. Identify deployable vertical outcomes, acceptance criteria and prerequisites; avoid one disconnected layer per milestone.
+2. Map task/repository/owner dependencies and integration order; expose critical blockers and external confirmations.
+3. Use team estimates and known capacity, record ranges/assumptions; when absent propose order, not invented dates or velocity.
+4. Define completion evidence and planned quality/security/architecture gates proportional to risk.
+5. Include migration/release/rollback responsibilities with DevOps, and feedback checkpoints with PO.
+6. Publish a plan draft for actual owner review; update status only from observed completed work and source links.
+
+## Output
+docs/delivery/delivery-plan.md linked to stories, ADRs, OpenSpec tasks and risks.
+
+## Template
+Read `.harness/capabilities/delivery/templates/delivery-plan.md` from the installed snapshot.
+Use `aih new-doc delivery-plan --capability delivery` only for a missing draft;
+reuse existing authoritative docs or a compact task note when appropriate. Never overwrite approved content.
+
+## Example
+Customer-search delivery: authorization + create + query + tests form the first usable slice. Bulk import follows schema readiness. With no capacity estimate, order the slices and label dates unknown.
+
+## Quality checklist
+- Every slice has user value and observable acceptance
+- Dependencies and accountable owners are explicit
+- Dates/capacity have sources and completion has evidence
+
+## Handoff
+State source revisions, requirement/decision IDs, approval status, unresolved questions, output paths,
+actual checks and next role. Use the module workflow; role names are not human authorization.
+
+## Constraints
+Follow Core permissions, clarification and project-conventions. Ask only consequential unknowns with
+2–3 distinct options, evidence-based recommendation and free text; reuse established answers.
+No invented source, measured result, target, estimate or approval. No external messaging/publishing,
+paid tool use or production changes beyond the user's explicit authorization.

package/capabilities/delivery/skills/harness-delivery-risk-management/SKILL.md

@@ -0,0 +1,42 @@
+---
+name: harness-delivery-risk-management
+description: Maintain owned delivery risks, assumptions, blockers and dependencies with response triggers.
+---
+
+# harness-delivery-risk-management
+
+## Input
+Plan, stakeholder/source constraints, technical/operational findings and actual progress evidence.
+
+## Procedure
+1. Identify risks versus current issues, assumptions and dependencies; link each to source and affected outcome.
+2. Assess impact and likelihood qualitatively or with actual evidence; do not invent numerical risk scores.
+3. Assign proposed/confirmed owner, mitigation, trigger, contingency and decision needed; clarify critical gaps.
+4. Track unresolved external dependencies and review cadence; no autonomous stakeholder messages or calendar writes.
+5. Update risk state only with evidence; carry accepted residual risks to release/PO decision and architecture review.
+
+## Output
+docs/delivery/risk-register.md and evidence-based status draft.
+
+## Template
+Read `.harness/capabilities/delivery/templates/risk-register.md` from the installed snapshot.
+Use `aih new-doc risk-register --capability delivery` only for a missing draft;
+reuse existing authoritative docs or a compact task note when appropriate. Never overwrite approved content.
+
+## Example
+Internal provider sandbox is unreachable: current blocker, not a completed test. Owner confirms access plan; integration milestone remains at risk until an observed connectivity check succeeds.
+
+## Quality checklist
+- Risk and observed issue are distinguished
+- Owner, trigger, response and source are present
+- Status does not fabricate progress or communicate externally
+
+## Handoff
+State source revisions, requirement/decision IDs, approval status, unresolved questions, output paths,
+actual checks and next role. Use the module workflow; role names are not human authorization.
+
+## Constraints
+Follow Core permissions, clarification and project-conventions. Ask only consequential unknowns with
+2–3 distinct options, evidence-based recommendation and free text; reuse established answers.
+No invented source, measured result, target, estimate or approval. No external messaging/publishing,
+paid tool use or production changes beyond the user's explicit authorization.

package/capabilities/delivery/skills/harness-delivery-scope-change/SKILL.md

@@ -0,0 +1,43 @@
+---
+name: harness-delivery-scope-change
+description: Evaluate scope changes across product value, architecture, dependencies and delivery before re-planning.
+---
+
+# harness-delivery-scope-change
+
+## Input
+Change request, reviewed baseline scope/MVP, affected requirements/ADRs, plan and available estimates.
+
+## Procedure
+1. Preserve the baseline and describe the proposed delta and motivation; identify decision owner.
+2. Ask BA/PO for requirement/value/AC impact, SA for contract/NFR/data impact and DevOps for release impact as applicable.
+3. Compare defer, replace and expand options with explicit trade-offs; avoid invented cost/dates.
+4. Resolve consequential owner decisions before implementing affected behavior; continue independent authorized work.
+5. Update only affected scope/traceability/ADR/plan records and OpenSpec tasks after the real decision.
+6. Report revised dependencies, acceptance, risks and rollback considerations; preserve history.
+
+## Output
+docs/delivery/change-assessment.md and reviewed updates to existing plan/requirements/design.
+
+## Template
+Read `.harness/capabilities/delivery/templates/change-assessment.md` from the installed snapshot.
+Use `aih new-doc change-assessment --capability delivery` only for a missing draft;
+reuse existing authoritative docs or a compact task note when appropriate. Never overwrite approved content.
+
+## Example
+Adding bulk import to the search MVP requires async jobs, limits and duplicate policy. Present defer or replace a lower-value feature; do not quietly promise the same release date.
+
+## Quality checklist
+- Baseline and delta are explicit and source-linked
+- Cross-role impacts and options are assessed
+- Scope/date approval is from the actual owner
+
+## Handoff
+State source revisions, requirement/decision IDs, approval status, unresolved questions, output paths,
+actual checks and next role. Use the module workflow; role names are not human authorization.
+
+## Constraints
+Follow Core permissions, clarification and project-conventions. Ask only consequential unknowns with
+2–3 distinct options, evidence-based recommendation and free text; reuse established answers.
+No invented source, measured result, target, estimate or approval. No external messaging/publishing,
+paid tool use or production changes beyond the user's explicit authorization.

package/capabilities/delivery/templates/change-assessment.md

@@ -0,0 +1,23 @@
+# Scope change assessment
+
+Status: draft / pending review
+Owner: TODO
+Sources and revisions: TODO
+Scope / related requirement and decision IDs: TODO
+Last verified against code: TODO (or not applicable with reason)
+
+## Baseline and requested delta
+Approved scope/version, request/source, motivation and actual decision owner.
+## Cross-role impact
+BA requirements/AC, PO value/MVP, SA NFR/module/data/contracts, PM dependency/estimate and DevOps release.
+## Options
+Defer, replace or expand where viable; value, risk, estimate confidence and rollback trade-offs.
+## Resolution and updates
+Actual owner answer/source; impacted docs, traceability, ADRs, tasks and plan updates after agreement.
+## Remaining uncertainty
+Unknown estimates/targets, external dependencies, blockers and independent work that can proceed.
+
+
+## Review and handoff
+Record actual owner decisions, unresolved consequential questions, verification performed/skipped,
+next role, source/output paths and affected OpenSpec change. Draft text is not an accepted decision.

package/capabilities/delivery/templates/delivery-plan.md

@@ -0,0 +1,25 @@
+# Delivery plan
+
+Status: draft / pending review
+Owner: TODO
+Sources and revisions: TODO
+Scope / related requirement and decision IDs: TODO
+Last verified against code: TODO (or not applicable with reason)
+
+## Baseline and authority
+Reviewed MVP/AC, architecture/source revisions, confirmed estimates/capacity or explicitly unknown.
+## Value slices and dependencies
+| Slice / user outcome | AC / OpenSpec tasks | Prerequisites / owner | Estimate source / range | Completion evidence |
+| --- | --- | --- | --- | --- |
+| TODO | TODO | TODO | unknown | pending |
+## Milestones
+Order and integration/repository merge sequence; dates are proposals with sources, not invented commitments.
+## Verification and release
+Quality gates, migration, deployment/rollback owners and feedback checkpoints; link DevOps plan.
+## Status and risks
+Actual completed evidence, blocked/pending work, risk-register links and next owner.
+
+
+## Review and handoff
+Record actual owner decisions, unresolved consequential questions, verification performed/skipped,
+next role, source/output paths and affected OpenSpec change. Draft text is not an accepted decision.

package/capabilities/delivery/templates/risk-register.md

@@ -0,0 +1,19 @@
+# Delivery risks and blockers
+
+Status: draft / pending review
+Owner: TODO
+Sources and revisions: TODO
+Scope / related requirement and decision IDs: TODO
+Last verified against code: TODO (or not applicable with reason)
+
+## Risk / issue / assumption / dependency register
+| ID / kind | Source / affected outcome | Impact / likelihood evidence | Owner status | Response / trigger / contingency | State / evidence |
+| --- | --- | --- | --- | --- | --- |
+| RISK-001 / TODO | TODO | unknown | proposed | TODO | pending |
+## Review cadence and decisions
+Actual confirmations, residual risk acceptances and decisions needed; no automatic stakeholder communication.
+
+
+## Review and handoff
+Record actual owner decisions, unresolved consequential questions, verification performed/skipped,
+next role, source/output paths and affected OpenSpec change. Draft text is not an accepted decision.

package/capabilities/delivery/workflows/execution-change.md

@@ -0,0 +1,9 @@
+# Progress, feedback and scope changes
+
+1. Read actual completed check/PR evidence against the current plan; do not infer completion from tasks alone.
+2. Maintain risk-register with source, confirmed/proposed owners, triggers and unresolved external dependencies.
+3. For changed scope load scope-change: preserve baseline, consult BA/PO/SA/DevOps as relevant, compare
+   defer/replace/expand and resolve material choices. Continue independent already-authorized work.
+4. Update affected plan, requirements, ADRs and OpenSpec only after the actual decision; keep history.
+5. Produce a status draft with completed/pending/blocked work and evidence. Release ownership/gates
+   remain in DevOps; PM proposals do not approve delivery or production changes.

package/capabilities/delivery/workflows/planning.md

@@ -0,0 +1,12 @@
+# Reviewed outcome → delivery plan
+
+Input: agreed scope/MVP, acceptance, solution boundaries and known team constraints.
+1. Read Core solution-framing and existing approved plan; choose compact or full depth.
+2. Load harness-delivery-planning for usable vertical slices, dependencies, repository order and gates.
+3. Use actual team estimates/capacity; without them document sequencing and missing estimate owners,
+   not fabricated milestones. Compare scope trade-offs with PO and technical dependencies with SA.
+4. Load risk-management for owned risks/blockers/assumptions/dependencies and response triggers.
+5. Link plan to requirements/ADRs/OpenSpec tasks and DevOps rollout/rollback. Actual owner reviews
+   consequential commitments; explicit authorization already covering routine tasks remains valid.
+6. Handoff approved slice, acceptance, prerequisites, verification, risks and next owner to Core/adapter.
+No automatic issue assignment, messaging, calendar entries, merge or deployment.

package/capabilities/product/README.md

@@ -76,3 +76,10 @@ After documented approval, UI/UX reads approved PRD, flows, stories and AC rathe
 ## Cross-capability walkthrough
 
 The complete path from source/idea to production is documented in [Full SDLC](../../docs/end-to-end-workflow.md). The [Skill matrix](../../docs/skill-matrix.md) maps all Product skills and outputs, and [Agent handoffs](../../docs/agent-handoffs.md) shows how approved requirements pass to UI/UX and OpenSpec. For an example, see [Full-stack web](../../docs/examples/fullstack-web.md).
+
+## MVP and solution handoff
+
+Use product-to-solution workflow, MVP/outcomes skills and their templates; load only relevant skills.
+Optional Architecture/Delivery provides SA/PM depth. Existing Product project-manager and delivery-plan
+remain compatibility entries; no automatic migration or stakeholder approval. All Product skills now
+include template/example/quality/handoff sections.

package/capabilities/product/agents/business-analyst.md

@@ -13,3 +13,6 @@ Read approved business sources only. Cannot invent stakeholder decisions or mark
 
 ## Handoff
 Hand reviewed requirements and source references to Product Manager and Product Owner; unresolved contradictions to decision owner.
+
+Pass source revisions, business rules, FR/US/AC/NFR IDs, contradictions and approval status to PO;
+use product-to-solution for accepted scope and return technical assumptions to Architecture.

package/capabilities/product/agents/product-owner.md

@@ -6,10 +6,14 @@ Owns proposed backlog structure, story readiness, acceptance coverage and priori
 Reviewed PRD, business rules, personas, constraints and team capacity signals.
 
 ## Outputs
-Epics, user stories, Gherkin acceptance criteria, backlog priority recommendations and release dependencies.
+Reviewed-scope proposals for complete MVP journeys, in/out scope, minimum quality, product outcome
+measures, epics/stories, Gherkin AC, priorities and release dependencies.
 
 ## Permissions and limits
 Do not self-approve or silently change stakeholder priorities; no invented sprint velocity or technical guarantees.
 
 ## Handoff
 Hand approved stories and traceable AC to UI/UX and OpenSpec; return gaps to BA/PM.
+
+Use product-to-solution workflow and MVP/outcomes skills; actual stakeholder decisions remain required
+for consequential scope choices. Send requirement/AC/NFR IDs and decision evidence to Architecture/Delivery.

package/capabilities/product/agents/project-manager.md

@@ -13,3 +13,8 @@ Schedules are proposals without team estimates and confirmations. No calendar/ev
 
 ## Handoff
 Send dependencies and risks to responsible owners, and scope-change impacts to Product Owner.
+
+Compatibility entry point: retain this role and the existing Product delivery-plan template for older
+consumers. For richer delivery planning, use the installed Delivery project-manager role/workflows
+when selected; otherwise keep sequencing/risks in existing project-owned docs. Do not auto-install it
+or migrate/overwrite the old plan. Project Manager manages delivery; Product Manager owns product value.

package/capabilities/product/capability.json

@@ -15,14 +15,17 @@
     "harness-product-idea-discovery",
     "harness-product-product-research",
     "harness-product-srs-generation",
-    "harness-product-use-case-generation"
+    "harness-product-use-case-generation",
+    "harness-product-mvp-definition",
+    "harness-product-product-outcomes"
   ],
   "workflows": [
     "workflows/idea-to-prd.md",
     "workflows/confluence-to-spec.md",
     "workflows/change-impact-analysis.md",
     "workflows/idea-driven.md",
-    "workflows/document-driven.md"
+    "workflows/document-driven.md",
+    "workflows/product-to-solution.md"
   ],
   "agents": [
     "agents/business-analyst.md",
@@ -47,9 +50,11 @@
     "templates/research.md",
     "templates/prd-lite.md",
     "templates/srs.md",
-    "templates/use-cases.md"
+    "templates/use-cases.md",
+    "templates/mvp.md",
+    "templates/outcomes.md"
   ],
-  "module_version": "1.1.0",
+  "module_version": "1.2.0",
   "verification": "checks/verification.json",
   "evaluation": "evals/cases.json"
 }

package/capabilities/product/checks/verification.json

@@ -36,6 +36,18 @@
         "OpenSpec handoff references approved scope only"
       ],
       "execution": "manual"
+    },
+    {
+      "id": "mvp-outcome",
+      "title": "MVP completeness and product outcome authority",
+      "command": [],
+      "when_file": null,
+      "evidence": [
+        "Usable end-to-end journey, explicit in/out scope and requirement/AC links",
+        "Minimum security/data integrity included",
+        "Metrics/targets and scope approval source or unresolved owner"
+      ],
+      "execution": "manual"
     }
   ]
 }

package/capabilities/product/evals/cases.json

@@ -30,6 +30,16 @@
         "Hands off only approved FR/NFR/US/AC and explicit source versions to OpenSpec"
       ],
       "critical": true
+    },
+    {
+      "id": "mvp-not-feature-count",
+      "prompt": "A user wants a customer search MVP, suggests skipping authorization and duplicate handling until later, and provides no approved success target.",
+      "criteria": [
+        "Defines a usable journey and explicit in/out scope with acceptance",
+        "Keeps minimum access control/data integrity and clarifies duplicate policy",
+        "Does not invent success metrics or approve the MVP on behalf of the owner"
+      ],
+      "critical": true
     }
   ]
 }