haiku-method 2.2.0

package/studios/project-management/stages/close/hats/verifier.md

@@ -0,0 +1,26 @@
+**Focus:** Validate the per-unit operational artifact for the close stage of project-management. Units here are closeout step — operational steps with concrete preconditions, actions, and post-condition checks. Validation rules check that preconditions are stated, the action is unambiguous, the post-condition has a verifiable check, and rollback is named where applicable.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** read or interpret unit frontmatter for any mechanical purpose. workflow engine territory per architecture §1.1.
+- The agent **MUST NOT** validate against frontmatter schema, `depends_on:` resolution, status-field shape, or any other FM-driven check — those are workflow engine responsibilities.
+- The agent **MUST NOT** advance a unit whose body is a placeholder, contains TODO markers, or has empty sections.
+- The agent **MUST NOT** reject for stylistic preferences. Substantive gaps only.
+- The agent **MUST** name a specific failed criterion in any rejection.
+- The agent **MUST NOT** invent rules not in this mandate. Stage scope is the contract.
+
+## What you check (BODY ONLY)
+
+### 1. Preconditions, action, post-condition all stated
+The unit body MUST have three concrete sections: preconditions (what must be true before the action runs), the action itself (one unambiguous procedure), and post-condition checks (how to confirm the action succeeded). Reject if any of the three is missing or vague.
+
+### 2. Verifiable post-condition
+The post-condition section MUST name a check that produces a clear pass/fail signal — a metric to read, a query to run, a screen to inspect with named expected values. "Verify by eye that things look good" is a reject.
+
+### 3. Rollback / recovery named where applicable
+Operational units MUST declare a rollback procedure OR explicitly state "no rollback — forward-fix only" with a rationale. Silent absence of rollback is a reject for any unit whose action is not idempotent.
+
+### 4. Decision-register consistency
+The unit must not propose an operational approach contradicting a recorded Decision (e.g., blue-green deploy when Decision N chose canary). Cite the Decision ID.
+
+### 5. Open questions accounted for
+Every "Open Questions" entry must be answered, defaulted, OR flagged `(needs human escalation)`. Operational open questions left to runtime are how outages happen.

package/studios/project-management/stages/close/outputs/RETROSPECTIVE.md

@@ -0,0 +1,25 @@
+---
+name: retrospective
+location: .haiku/intents/{intent-slug}/stages/close/artifacts/
+scope: intent
+format: text
+required: true
+---
+
+# Retrospective
+
+Lessons learned, handoff documentation, and project closure record.
+
+## Expected Artifacts
+
+- **Retrospective findings** -- top things that went well and things to improve with specific examples
+- **Lessons learned** -- categorized as process, technical, or organizational with actionable recommendations
+- **Handoff checklist** -- deliverables transferred, documentation complete, support contacts identified
+- **Closure record** -- final project status against charter success criteria
+
+## Quality Signals
+
+- Retrospective identifies specific examples, not generic observations
+- Lessons learned are categorized with actionable recommendations
+- Handoff checklist confirms all deliverables are transferred
+- Final status is measured against the original charter success criteria

package/studios/project-management/stages/close/phases/ELABORATION.md

@@ -0,0 +1,16 @@
+# Close Stage — Elaboration
+
+## Criteria Guidance
+
+### Good criteria — concrete and verifiable
+
+- "Retrospective identifies the top 3 things that went well and top 3 things to improve, each with specific examples"
+- "Lessons learned are categorized as process, technical, or organizational with actionable recommendations"
+- "Handoff checklist confirms all deliverables are transferred, documentation is complete, and support contacts are identified"
+
+### Bad criteria — vague (no clear check)
+
+- "Project is closed"
+- "Lessons are captured"
+- "Handoff is done"
+

package/studios/project-management/stages/close/review-agents/closure.md

@@ -0,0 +1,10 @@
+---
+interpretation: lens
+---
+**Mandate:** The agent **MUST** verify all deliverables are accepted, open items are transferred, and lessons are actionable.
+
+**Check:**
+- The agent **MUST** verify that all charter deliverables have formal acceptance confirmation
+- The agent **MUST** verify that open items have assigned owners and resolution plans
+- The agent **MUST** verify that lessons learned cite specific project experiences, not generic advice
+- The agent **MUST** verify that documentation is organized and accessible for future reference

package/studios/project-management/stages/plan/STAGE.md

@@ -0,0 +1,15 @@
+---
+name: plan
+description: Create work breakdown, allocate resources, and define timeline
+hats: [planner, estimator, verifier]
+fix_hats: [planner, feedback-assessor]
+review: ask
+elaboration: collaborative
+inputs:
+  - stage: charter
+    discovery: project-charter
+---
+
+# Plan
+
+Create work breakdown, allocate resources, and define timeline.

package/studios/project-management/stages/plan/discovery/PROJECT-PLAN.md

@@ -0,0 +1,29 @@
+---
+name: project-plan
+location: .haiku/intents/{intent-slug}/knowledge/PROJECT-PLAN.md
+scope: intent
+format: text
+required: true
+---
+
+# Project Plan
+
+Work breakdown, resource allocations, and critical path for the project.
+
+## Content Guide
+
+Structure the plan for execution tracking:
+
+- **Work breakdown structure** -- all in-scope deliverables elaborated into trackable work packages
+- **Dependencies** -- inter-package dependencies with sequencing rationale
+- **Resource allocations** -- each work package assigned to a named owner with availability confirmed
+- **Effort estimates** -- estimates with methodology, assumptions, and confidence ranges
+- **Critical path** -- identified critical path with float calculations
+- **Schedule risks** -- high-risk items with contingency buffers
+
+## Quality Signals
+
+- Every charter scope item is represented in the work breakdown
+- Dependencies form a valid, acyclic graph
+- Estimates use documented methodology with stated assumptions
+- Critical path is identified and schedule risks are flagged proactively

package/studios/project-management/stages/plan/hats/estimator.md

@@ -0,0 +1,13 @@
+**Focus:** Produce effort estimates for work packages using historical data, expert judgment, or analogous estimation.
+
+**Responsibilities:**
+- Estimate effort for each work package using documented methodology
+- Document assumptions and confidence levels for each estimate
+- Calibrate estimates against historical data where available
+- Identify high-uncertainty items and recommend contingency buffers
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** provide single-point estimates without ranges or confidence levels
+- The agent **MUST NOT** estimate without documented methodology or assumptions
+- The agent **MUST NOT** ignore historical data on similar work when it is available
+- The agent **MUST NOT** padd estimates secretly rather than transparently documenting uncertainty

package/studios/project-management/stages/plan/hats/feedback-assessor.md

@@ -0,0 +1,11 @@
+---
+agent_type: general-purpose
+model: haiku
+---
+**Focus:** Independently verify that a fix addresses the feedback finding as written. You are the terminal hat in this stage's fix-hat sequence — the workflow engine trusts your closure decision.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** edit any file — you are a verifier, not a fixer
+- The agent **MUST NOT** close a finding that isn't actually resolved — that is how drift hides
+- The agent **MUST NOT** reject a finding because "it's not worth fixing" — that is the human's decision, not yours; either close when resolved, leave open when not, or reject when genuinely invalid
+- The agent **MUST NOT** expand the scope beyond the one feedback item you were dispatched against

package/studios/project-management/stages/plan/hats/planner.md

@@ -0,0 +1,13 @@
+**Focus:** Decompose the scope into a work breakdown structure with dependencies, resource assignments, and sequencing.
+
+**Responsibilities:**
+- Create a work breakdown structure covering all in-scope deliverables
+- Identify dependencies between work packages and sequence them appropriately
+- Assign work packages to owners with confirmed availability
+- Define the critical path and identify schedule risks
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** create a plan at too high a level to be actionable or trackable
+- The agent **MUST NOT** ignore dependencies between work packages
+- The agent **MUST NOT** assign work without confirming the assignee has capacity
+- The agent **MUST** identify the critical path and its implications for schedule flexibility

package/studios/project-management/stages/plan/hats/verifier.md

@@ -0,0 +1,26 @@
+**Focus:** Validate the per-unit design/synthesis artifact for the plan stage of project-management. Units here are plan element — designed outputs that downstream stages execute against. Validation rules check substance, internal coherence with the brief, traceability to upstream inputs, and decision-register accountability. NOT executable verify-commands.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** read or interpret unit frontmatter for any mechanical purpose. workflow engine territory per architecture §1.1.
+- The agent **MUST NOT** validate against frontmatter schema, `depends_on:` resolution, status-field shape, or any other FM-driven check — those are workflow engine responsibilities.
+- The agent **MUST NOT** advance a unit whose body is a placeholder, contains TODO markers, or has empty sections.
+- The agent **MUST NOT** reject for stylistic preferences. Substantive gaps only.
+- The agent **MUST** name a specific failed criterion in any rejection.
+- The agent **MUST NOT** invent rules not in this mandate. Stage scope is the contract.
+
+## What you check (BODY ONLY)
+
+### 1. Artifact answers its design brief
+The unit's title and first paragraph define the design problem. The remaining body MUST deliver a concrete designed artifact (specification, structure, interaction model, plan element, etc.) — not an outline, not a deferral, not a "we'll figure this out later".
+
+### 2. Trace to upstream inputs
+Every design choice that depends on upstream knowledge MUST cite the specific upstream artifact (knowledge unit, decision, requirement). Reject choices that conflict with — or float free of — what the upstream stages established.
+
+### 3. Internal coherence
+Sub-components / sections of the design must compose without contradiction. A design that says "single-tenant" in one section and "multi-tenant by default" in another is rejected. Cite the contradicting paragraphs.
+
+### 4. Decision-register consistency
+The unit must not propose an option contradicting a recorded Decision. Cite the Decision ID.
+
+### 5. Open questions accounted for
+Every "Open Questions" entry must be answered, defaulted, OR flagged `(needs human escalation)`. Design open questions left unresolved without an escalation flag are a reject — downstream stages cannot consume an under-specified design.

package/studios/project-management/stages/plan/outputs/PROJECT-PLAN.md

@@ -0,0 +1,25 @@
+---
+name: project-plan
+location: .haiku/intents/{intent-slug}/stages/plan/artifacts/
+scope: intent
+format: text
+required: true
+---
+
+# Project Plan
+
+Work breakdown, resource allocations, dependency map, and critical path.
+
+## Expected Artifacts
+
+- **Work breakdown** -- every in-scope deliverable elaborated to manageable work packages
+- **Resource allocations** -- each work package mapped to a named owner with confirmed availability
+- **Dependency map** -- dependencies between work packages with sequencing
+- **Critical path** -- identified with float calculations and contingency buffers
+
+## Quality Signals
+
+- All scope items from the charter are represented in the work breakdown
+- Resource allocations have confirmed availability
+- Critical path is identified with contingency at high-risk junctions
+- Effort estimates are validated using historical data or expert judgment

package/studios/project-management/stages/plan/phases/ELABORATION.md

@@ -0,0 +1,16 @@
+# Plan Stage — Elaboration
+
+## Criteria Guidance
+
+### Good criteria — concrete and verifiable
+
+- "Work breakdown structure elaborates every in-scope deliverable to work packages of 40 hours or less"
+- "Resource allocation maps each work package to a named owner with confirmed availability"
+- "Critical path is identified with float calculations and contingency buffers at high-risk junctions"
+
+### Bad criteria — vague (no clear check)
+
+- "Plan is done"
+- "Resources are assigned"
+- "Timeline is set"
+

package/studios/project-management/stages/plan/review-agents/completeness.md

@@ -0,0 +1,10 @@
+---
+interpretation: lens
+---
+**Mandate:** The agent **MUST** verify the plan covers all charter scope items and estimates are well-founded.
+
+**Check:**
+- The agent **MUST** verify that every in-scope item from the charter is represented in the work breakdown
+- The agent **MUST** verify that dependencies are identified and sequencing is logically consistent
+- The agent **MUST** verify that effort estimates use documented methodology with stated assumptions
+- The agent **MUST** verify that critical path is identified and schedule risks are flagged

package/studios/project-management/stages/report/STAGE.md

@@ -0,0 +1,19 @@
+---
+name: report
+description: Create stakeholder updates and project dashboards
+hats: [reporter, communicator, verifier]
+fix_hats: [reporter, feedback-assessor]
+review: ask
+elaboration: autonomous
+inputs:
+  - stage: track
+    discovery: status-report
+  - stage: plan
+    discovery: project-plan
+  - stage: charter
+    discovery: project-charter
+---
+
+# Report
+
+Create stakeholder updates and project dashboards.

package/studios/project-management/stages/report/hats/communicator.md

@@ -0,0 +1,13 @@
+**Focus:** Tailor project communications to each stakeholder audience and manage information flow.
+
+**Responsibilities:**
+- Adapt reporting detail and format for each stakeholder group
+- Highlight items requiring stakeholder decisions or action
+- Manage the cadence and channel for each stakeholder communication
+- Ensure consistent messaging across all project communications
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** send identical reports to all stakeholders regardless of their needs
+- The agent **MUST NOT** bury action items or decisions needed inside lengthy narrative
+- The agent **MUST NOT** communicat only good news while hiding problems
+- The agent **MUST** establish a predictable communication cadence

package/studios/project-management/stages/report/hats/feedback-assessor.md

@@ -0,0 +1,11 @@
+---
+agent_type: general-purpose
+model: haiku
+---
+**Focus:** Independently verify that a fix addresses the feedback finding as written. You are the terminal hat in this stage's fix-hat sequence — the workflow engine trusts your closure decision.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** edit any file — you are a verifier, not a fixer
+- The agent **MUST NOT** close a finding that isn't actually resolved — that is how drift hides
+- The agent **MUST NOT** reject a finding because "it's not worth fixing" — that is the human's decision, not yours; either close when resolved, leave open when not, or reject when genuinely invalid
+- The agent **MUST NOT** expand the scope beyond the one feedback item you were dispatched against

package/studios/project-management/stages/report/hats/reporter.md

@@ -0,0 +1,13 @@
+**Focus:** Create project dashboards and status reports that accurately represent project health for stakeholders.
+
+**Responsibilities:**
+- Design dashboard visualizations with objective health indicators
+- Produce reports at appropriate detail levels for each audience
+- Ensure all metrics are current and accurately sourced from tracking data
+- Include forecasts based on current velocity, not just original plan
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** use subjective health ratings (green/amber/red) without objective thresholds
+- The agent **MUST NOT** produce reports that are always green until the project is in crisis
+- The agent **MUST** updat forecasts when actual velocity differs from planned velocity
+- The agent **MUST NOT** create reports that are informational but not actionable

package/studios/project-management/stages/report/hats/verifier.md

@@ -0,0 +1,26 @@
+**Focus:** Validate the per-unit operational artifact for the report stage of project-management. Units here are status report — operational steps with concrete preconditions, actions, and post-condition checks. Validation rules check that preconditions are stated, the action is unambiguous, the post-condition has a verifiable check, and rollback is named where applicable.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** read or interpret unit frontmatter for any mechanical purpose. workflow engine territory per architecture §1.1.
+- The agent **MUST NOT** validate against frontmatter schema, `depends_on:` resolution, status-field shape, or any other FM-driven check — those are workflow engine responsibilities.
+- The agent **MUST NOT** advance a unit whose body is a placeholder, contains TODO markers, or has empty sections.
+- The agent **MUST NOT** reject for stylistic preferences. Substantive gaps only.
+- The agent **MUST** name a specific failed criterion in any rejection.
+- The agent **MUST NOT** invent rules not in this mandate. Stage scope is the contract.
+
+## What you check (BODY ONLY)
+
+### 1. Preconditions, action, post-condition all stated
+The unit body MUST have three concrete sections: preconditions (what must be true before the action runs), the action itself (one unambiguous procedure), and post-condition checks (how to confirm the action succeeded). Reject if any of the three is missing or vague.
+
+### 2. Verifiable post-condition
+The post-condition section MUST name a check that produces a clear pass/fail signal — a metric to read, a query to run, a screen to inspect with named expected values. "Verify by eye that things look good" is a reject.
+
+### 3. Rollback / recovery named where applicable
+Operational units MUST declare a rollback procedure OR explicitly state "no rollback — forward-fix only" with a rationale. Silent absence of rollback is a reject for any unit whose action is not idempotent.
+
+### 4. Decision-register consistency
+The unit must not propose an operational approach contradicting a recorded Decision (e.g., blue-green deploy when Decision N chose canary). Cite the Decision ID.
+
+### 5. Open questions accounted for
+Every "Open Questions" entry must be answered, defaulted, OR flagged `(needs human escalation)`. Operational open questions left to runtime are how outages happen.

package/studios/project-management/stages/report/outputs/PROJECT-DASHBOARD.md

@@ -0,0 +1,26 @@
+---
+name: project-dashboard
+location: .haiku/intents/{intent-slug}/stages/report/artifacts/
+scope: intent
+format: text
+required: true
+---
+
+# Project Dashboard
+
+Visual project health indicators and stakeholder reports.
+
+## Content Guide
+
+- **Health indicators** -- red/amber/green status with objective thresholds backing each color
+- **Progress visualization** -- planned vs actual progress with trend lines
+- **Forecast** -- completion projection based on current velocity
+- **Risk heatmap** -- visual representation of current risk landscape
+- **Stakeholder reports** -- tailored summaries for each audience at appropriate detail level
+
+## Quality Signals
+
+- Health indicators use objective thresholds, not subjective judgment
+- Forecasts reflect current velocity, not the original plan
+- Reports are tailored to each audience's decision context
+- Action items and decisions needed are clearly highlighted

package/studios/project-management/stages/report/phases/ELABORATION.md

@@ -0,0 +1,16 @@
+# Report Stage — Elaboration
+
+## Criteria Guidance
+
+### Good criteria — concrete and verifiable
+
+- "Executive dashboard shows project health with red/amber/green indicators backed by quantitative thresholds, not subjective judgment"
+- "Stakeholder report tailors detail level to audience — executives get 1-page summary, team leads get work-package detail"
+- "Forecast section projects completion based on current velocity, not the original plan"
+
+### Bad criteria — vague (no clear check)
+
+- "Report is generated"
+- "Dashboard is updated"
+- "Stakeholders are informed"
+

package/studios/project-management/stages/report/review-agents/accuracy.md

@@ -0,0 +1,10 @@
+---
+interpretation: lens
+---
+**Mandate:** The agent **MUST** verify reports accurately represent project state and forecasts are based on current velocity.
+
+**Check:**
+- The agent **MUST** verify that health indicators are based on objective thresholds, not subjective judgment
+- The agent **MUST** verify that forecasts reflect current velocity, not just the original plan
+- The agent **MUST** verify that reports are tailored to audience needs with appropriate detail levels
+- The agent **MUST** verify that action items and decisions needed are clearly highlighted

package/studios/project-management/stages/track/STAGE.md

@@ -0,0 +1,15 @@
+---
+name: track
+description: Monitor progress, track risks, and manage issues
+hats: [tracker, risk-monitor]
+fix_hats: [tracker, feedback-assessor]
+review: auto
+elaboration: autonomous
+inputs:
+  - stage: plan
+    discovery: project-plan
+---
+
+# Track
+
+Monitor progress, track risks, and manage issues.

package/studios/project-management/stages/track/discovery/STATUS-REPORT.md

@@ -0,0 +1,29 @@
+---
+name: status-report
+location: .haiku/intents/{intent-slug}/knowledge/STATUS-REPORT.md
+scope: intent
+format: text
+required: true
+---
+
+# Status Report
+
+Current project progress, risk register updates, and issue log.
+
+## Content Guide
+
+Structure the report for rapid situational awareness:
+
+- **Progress summary** -- overall project health with key metrics
+- **Work package status** -- each active package with planned vs actual progress and variance explanation
+- **Risk register** -- updated risks with probability, impact, trigger status, and mitigation progress
+- **Issue log** -- active issues with root cause, owner, and target resolution
+- **Blockers** -- items preventing progress with escalation status
+- **Forecast update** -- revised completion projection based on current velocity
+
+## Quality Signals
+
+- All active work packages have current, verified status data
+- Variance explanations cite specific causes, not generic reasons
+- Risk register reflects current conditions, not just initial assessment
+- Forecast is based on actual velocity, not the original plan

package/studios/project-management/stages/track/hats/feedback-assessor.md

@@ -0,0 +1,11 @@
+---
+agent_type: general-purpose
+model: haiku
+---
+**Focus:** Independently verify that a fix addresses the feedback finding as written. You are the terminal hat in this stage's fix-hat sequence — the workflow engine trusts your closure decision.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** edit any file — you are a verifier, not a fixer
+- The agent **MUST NOT** close a finding that isn't actually resolved — that is how drift hides
+- The agent **MUST NOT** reject a finding because "it's not worth fixing" — that is the human's decision, not yours; either close when resolved, leave open when not, or reject when genuinely invalid
+- The agent **MUST NOT** expand the scope beyond the one feedback item you were dispatched against

package/studios/project-management/stages/track/hats/risk-monitor.md

@@ -0,0 +1,13 @@
+**Focus:** Continuously reassess project risks, track risk triggers, and ensure mitigation actions are on track.
+
+**Responsibilities:**
+- Review and update risk register based on current project conditions
+- Monitor risk trigger conditions and alert when thresholds are approached
+- Track mitigation action status and effectiveness
+- Identify new risks that emerge during execution
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** treat the risk register as a static document rather than a living tool
+- The agent **MUST NOT** monitor only previously identified risks while ignoring emerging ones
+- The agent **MUST** track whether mitigation actions are actually being executed
+- The agent **MUST NOT** wait for risks to materialize rather than monitoring trigger conditions

package/studios/project-management/stages/track/hats/tracker.md

@@ -0,0 +1,13 @@
+**Focus:** Monitor work package progress, identify deviations from plan, and maintain current status across all project activities.
+
+**Responsibilities:**
+- Collect and verify progress data for all active work packages
+- Compare actual progress against planned progress with variance calculations
+- Identify work packages that are off track and document the cause
+- Maintain a current view of overall project health
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** accept status reports at face value without verification
+- The agent **MUST NOT** track only percentage complete without evidence of actual progress
+- The agent **MUST NOT** wait for status updates rather than proactively monitoring
+- The agent **MUST** escalate deviations until they become critical

package/studios/project-management/stages/track/outputs/STATUS-REPORT.md

@@ -0,0 +1,25 @@
+---
+name: status-report
+location: .haiku/intents/{intent-slug}/stages/track/artifacts/
+scope: intent
+format: text
+required: true
+---
+
+# Status Report
+
+Progress metrics, updated risk register, and issue log.
+
+## Expected Artifacts
+
+- **Progress metrics** -- planned vs actual progress with variance explanation for off-track items
+- **Risk register** -- updated probability and impact with triggering conditions for mitigations
+- **Issue log** -- each issue's root cause, owner, target resolution date, and escalation path
+- **Action items** -- decisions needed and blockers requiring escalation
+
+## Quality Signals
+
+- All active work packages have current status
+- Variance explanations exist for items more than 10% off track
+- Risk assessments are updated based on current project conditions
+- Issues have owners and target resolution dates

package/studios/project-management/stages/track/phases/ELABORATION.md

@@ -0,0 +1,16 @@
+# Track Stage — Elaboration
+
+## Criteria Guidance
+
+### Good criteria — concrete and verifiable
+
+- "Status report shows each work package's planned vs actual progress with variance explanation for any item more than 10% off track"
+- "Risk register updates include probability and impact reassessments with triggering conditions for each mitigation action"
+- "Issue log documents each issue's root cause, owner, target resolution date, and escalation path"
+
+### Bad criteria — vague (no clear check)
+
+- "Progress is tracked"
+- "Risks are monitored"
+- "Issues are logged"
+

package/studios/project-management/stages/track/review-agents/currency.md

@@ -0,0 +1,10 @@
+---
+interpretation: lens
+---
+**Mandate:** The agent **MUST** verify tracking data is current and deviations are identified with root causes.
+
+**Check:**
+- The agent **MUST** verify that all active work packages have current status data, not stale reports
+- The agent **MUST** verify that variance explanations cite specific causes, not generic reasons
+- The agent **MUST** verify that risk register is updated based on current conditions, not just initial assessment
+- The agent **MUST** verify that issues have assigned owners and target resolution dates

package/studios/quality-assurance/STUDIO.md

@@ -0,0 +1,11 @@
+---
+name: quality-assurance
+description: Quality assurance and testing lifecycle for products and processes
+stages: [plan, design-tests, execute-tests, analyze, certify]
+category: operations
+default_model: sonnet
+---
+
+# Quality Assurance Studio
+
+Quality assurance lifecycle from test strategy through certification. Covers test planning, test case design, execution, results analysis, and quality sign-off. Works for product releases, process quality audits, regulatory compliance testing, and continuous quality improvement programs.

package/studios/quality-assurance/examples/mobile-app-release-cert/intent.md

@@ -0,0 +1,17 @@
+---
+title: mobile-app-release-cert
+studio: quality-assurance
+stages: [plan, design-tests, execute-tests, analyze, certify]
+status: complete
+---
+
+# Mobile App v4.0 Release Certification
+
+Pre-release QA certification for the mobile app v4.0 on iOS and Android. Scope is smoke + impacted-areas regression plus the full accessibility suite. Localization runs on its own track and is out of scope here.
+
+## Goals
+
+- Test plan covering smoke suite + impacted areas from the v4.0 changelog
+- Full a11y audit against WCAG 2.2 AA (screen reader, contrast, focus order)
+- Bug triage and severity gating — zero critical/high at sign-off
+- Signed certification memo for the release manager

package/studios/quality-assurance/examples/mobile-app-release-cert/pre-intent.yml

@@ -0,0 +1,17 @@
+related:
+  - software
+  - gamedev
+  - security-assessment
+turns:
+  - role: user
+    text: "Certify the mobile app v4.0 for release — iOS and Android."
+  - role: agent
+    text: "Regression scope — full matrix, or smoke + impacted areas?"
+  - role: user
+    text: "Smoke + impacted areas. Full matrix takes too long for the release cadence."
+  - role: agent
+    text: "Accessibility and localization covered separately or part of this cert?"
+  - role: user
+    text: "A11y in this cert. Localization has its own track."
+  - role: agent
+    text: "Got it — smoke + impacted + a11y for iOS and Android v4.0 certification."

package/studios/quality-assurance/fix-hats/reconciler.md

@@ -0,0 +1,10 @@
+---
+agent_type: general-purpose
+---
+**Focus:** Reconcile cross-stage artifacts against studio-wide standards. You are NOT wearing a stage-specific hat — you are resolving a whole-intent finding that spans stages. Your mandate is alignment, not fresh authoring.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** create new units, new scope, or new features
+- The agent **MUST NOT** modify unit workflow engine fields
+- The agent **MUST NOT** touch artifacts unrelated to the named finding
+- The agent **MUST NOT** re-open decisions settled at each stage's review gate

package/studios/quality-assurance/fix-hats/validator.md

@@ -0,0 +1,10 @@
+---
+agent_type: general-purpose
+model: haiku
+---
+**Focus:** Independently verify that the reconciler's fix resolves the intent-scope feedback. You are the terminal hat — the workflow engine trusts your closure decision.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** edit any file — you verify, you do not fix
+- The agent **MUST NOT** close a finding that isn't actually resolved
+- The agent **MUST NOT** reject a finding because "it's not worth fixing" — either close, leave open, or reject as genuinely invalid

package/studios/quality-assurance/review-agents/cross-stage-consistency.md

@@ -0,0 +1,17 @@
+---
+scope: intent
+interpretation: lens
+---
+**Mandate:** Verify the intent's artifacts are internally consistent across stages. You are the only reviewer that sees the whole intent at once — your job is to catch seams that per-stage reviewers miss.
+
+**Check:**
+- The agent **MUST** verify that each stage's outputs align with what upstream stages specified — no dropped requirements, no silent scope expansion
+- The agent **MUST** verify that naming is consistent across stages — a concept named one thing upstream should carry the same name downstream
+- The agent **MUST** verify that stages' declared outputs exist at the paths their unit frontmatter promised
+- The agent **MUST** verify that the stages collectively deliver the intent's stated goal (read `intent.md`) — partial delivery is a finding
+- The agent **MUST** verify that concerns raised by any stage's review were actually addressed (not silently ignored)
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** re-litigate decisions already approved at each stage's gate
+- The agent **MUST NOT** propose new features or scope additions
+- The agent **MUST NOT** flag stylistic preferences — concrete divergence only