haiku-method 2.2.0

package/studios/product-strategy/operations/user-signal-review.md

@@ -0,0 +1,20 @@
+---
+name: user-signal-review
+type: scheduled
+owner: human
+frequency: "biweekly"
+---
+
+**Purpose:** Aggregate and review user feedback signals to inform prioritization. Anecdotes aggregate into patterns.
+
+**Procedure:**
+- [ ] Review support tickets tagged as feature requests
+- [ ] Check NPS verbatim responses for product themes
+- [ ] Review sales objections related to missing features
+- [ ] Aggregate signals by theme and frequency
+- [ ] Update opportunity scores if patterns shift
+
+**Signals:**
+- Biweekly cadence
+- Spike in a particular feature request
+- Sales reports recurring objection pattern

package/studios/product-strategy/reflections/market-fit.md

@@ -0,0 +1,17 @@
+---
+name: market-fit
+studio: product-strategy
+---
+
+**Analyze:** User research predictions vs actual user behavior, competitive positioning accuracy.
+
+**Look for:**
+- User needs that research identified but the roadmap didn't address
+- Competitive threats that materialized vs were overestimated
+- Whether user segments behaved as the research predicted
+- Market assumptions that proved wrong
+
+**Produce:**
+- Research-to-reality comparison
+- Competitive positioning accuracy assessment
+- Recommendations for research methodology improvements

package/studios/product-strategy/reflections/prioritization-accuracy.md

@@ -0,0 +1,17 @@
+---
+name: prioritization-accuracy
+studio: product-strategy
+---
+
+**Analyze:** Did the highest-priority items deliver the predicted impact? Were deprioritized items correctly deprioritized?
+
+**Look for:**
+- Items that were prioritized high but delivered low impact (overestimated)
+- Items that were deprioritized but later became urgent (underestimated)
+- Whether the scoring framework produced defensible rankings
+- Stakeholder overrides that improved or worsened outcomes
+
+**Produce:**
+- Prioritization accuracy assessment
+- Scoring framework calibration recommendations
+- Recommendations for reducing stakeholder bias in prioritization

package/studios/product-strategy/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

package/studios/product-strategy/stages/discovery/STAGE.md

@@ -0,0 +1,13 @@
+---
+name: discovery
+description: Explore market landscape, competitive positioning, and opportunity space
+hats: [market-explorer, competitive-analyst, verifier]
+fix_hats: [market-explorer, feedback-assessor]
+review: auto
+elaboration: autonomous
+inputs: []
+---
+
+# Discovery
+
+Explore market landscape, competitive positioning, and opportunity space.

package/studios/product-strategy/stages/discovery/discovery/MARKET-LANDSCAPE.md

@@ -0,0 +1,29 @@
+---
+name: market-landscape
+location: .haiku/intents/{intent-slug}/knowledge/MARKET-LANDSCAPE.md
+scope: intent
+format: text
+required: true
+---
+
+# Market Landscape
+
+Document the market landscape, competitive positioning, and opportunity space. This output feeds downstream stages (user-research, prioritization, roadmap) as their foundational market context.
+
+## Content Guide
+
+Structure the landscape around the strategic questions the intent needs answered:
+
+- **Market segments** — identified segments with size estimates, growth indicators, and key dynamics
+- **Competitive positioning** — direct competitors, substitutes, and their strategic positions with feature and go-to-market comparisons
+- **Positioning map** — visual or structured representation of where players sit relative to key dimensions
+- **Emerging trends** — market shifts, technology changes, or regulatory developments that create or close opportunities
+- **Opportunity signals** — underserved segments, positioning gaps, or unmet needs surfaced by the analysis
+- **Sources** — where market data came from, with retrieval dates
+
+## Quality Signals
+
+- Market claims cite specific sources, not general industry knowledge
+- Competitive analysis covers positioning strategy, not just feature lists
+- Opportunity signals are evidence-backed, not speculative
+- The landscape is organized for consumption by the user-research stage, not as a raw data dump

package/studios/product-strategy/stages/discovery/hats/competitive-analyst.md

@@ -0,0 +1,8 @@
+**Focus:** Analyze competitors' positioning, strengths, weaknesses, and strategic direction. Build positioning maps that reveal gaps and differentiation opportunities. Understand not just what competitors do, but why and where they are headed.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** list competitors without analyzing their strategic positioning
+- The agent **MUST NOT** treat competitor feature lists as the full picture while ignoring go-to-market strategy
+- The agent **MUST NOT** only analyze direct competitors while ignoring substitutes and alternatives
+- The agent **MUST NOT** assume current positioning is static — missing competitor trajectory
+- The agent **MUST NOT** let competitor analysis anchor thinking instead of revealing open space

package/studios/product-strategy/stages/discovery/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/product-strategy/stages/discovery/hats/market-explorer.md

@@ -0,0 +1,8 @@
+**Focus:** Scan the market landscape — segments, trends, emerging technologies, and adjacent spaces. Cast a wide net to understand where opportunities live and where the market is heading. Breadth over depth at this stage.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** fixat on a single market segment before surveying the full landscape
+- The agent **MUST NOT** rely solely on incumbent analysis while missing emerging entrants
+- The agent **MUST NOT** report trends without citing sources or evidence
+- The agent **MUST NOT** confus market noise with genuine signals
+- The agent **MUST NOT** ignore adjacent markets that could expand or threaten the opportunity space

package/studios/product-strategy/stages/discovery/hats/verifier.md

@@ -0,0 +1,26 @@
+**Focus:** Validate the per-unit knowledge artifact for the discovery stage of product-strategy. Units here are market/user insight — knowledge artifacts that downstream stages consume. Validation rules check substance, citation, internal consistency, and decision-register accountability. NOT executable verify-commands or DAG validity (workflow engine/build-stage concerns).
+
+**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 topic
+The unit's title and first paragraph define the topic. The remaining body MUST deliver substantive content on that topic. Reject placeholders, content-free outlines, or redirects.
+
+### 2. Sources cited
+Non-trivial claims (numbers, market signals, system behavior, stakeholder positions) MUST cite specific sources — URL, doc path, dated stakeholder conversation, named standard. Reject "industry common knowledge" or unsourced numerical claims.
+
+### 3. Internal consistency
+Title, mission, and body must align. Numerical/categorical claims must be consistent across the body. Recommendations must follow from the evidence presented.
+
+### 4. Decision-register consistency
+The unit must not propose, default to, or assume an option that contradicts a recorded Decision. Cite the Decision ID in any rejection.
+
+### 5. Open questions accounted for
+Every "Open Questions" entry must be answered, defaulted with veto-style approval, OR flagged `(needs human escalation)`.

package/studios/product-strategy/stages/discovery/outputs/MARKET-LANDSCAPE.md

@@ -0,0 +1,25 @@
+---
+name: market-landscape
+location: .haiku/intents/{intent-slug}/stages/discovery/artifacts/
+scope: intent
+format: text
+required: true
+---
+
+# Market Landscape
+
+Market segment analysis, competitive positioning map, and opportunity space assessment.
+
+## Expected Artifacts
+
+- **Segment analysis** -- at least 3 market segments with size estimates and growth trends
+- **Competitive positioning map** -- at least 5 direct competitors with positioning analysis
+- **Opportunity space** -- underserved segments with evidence-backed sizing
+- **Differentiation vectors** -- positioning gaps and potential differentiation opportunities
+
+## Quality Signals
+
+- All market claims cite sources with dates
+- Competitive analysis identifies positioning gaps
+- Opportunity space is mapped with evidence-backed sizing
+- At least 3 underserved segments are identified with evidence

package/studios/product-strategy/stages/discovery/phases/ELABORATION.md

@@ -0,0 +1,16 @@
+# Discovery Stage — Elaboration
+
+## Criteria Guidance
+
+### Good criteria — concrete and verifiable
+
+- "Market landscape covers at least 3 segments with size estimates and growth trends"
+- "Competitive analysis identifies at least 5 direct competitors with positioning maps"
+- "Opportunity space includes at least 3 underserved segments with evidence for each"
+
+### Bad criteria — vague (no clear check)
+
+- "Market research is complete"
+- "Competitors are listed"
+- "Opportunities are identified"
+

package/studios/product-strategy/stages/discovery/review-agents/thoroughness.md

@@ -0,0 +1,10 @@
+---
+interpretation: lens
+---
+**Mandate:** The agent **MUST** verify the market landscape analysis is comprehensive and evidence-based.
+
+**Check:**
+- The agent **MUST** verify that competitive analysis covers direct competitors and adjacent solutions
+- The agent **MUST** verify that market sizing is based on verifiable data, not top-down guesses
+- The agent **MUST** verify that opportunity identification is grounded in observed user behavior or stated needs
+- The agent **MUST** verify that assumptions are labeled as such and have validation plans

package/studios/product-strategy/stages/prioritization/STAGE.md

@@ -0,0 +1,15 @@
+---
+name: prioritization
+description: Score and rank opportunities using impact/effort frameworks
+hats: [prioritizer, stakeholder-proxy, verifier]
+fix_hats: [prioritizer, feedback-assessor]
+review: ask
+elaboration: collaborative
+inputs:
+  - stage: user-research
+    discovery: insights-report
+---
+
+# Prioritization
+
+Score and rank opportunities using impact/effort frameworks.

package/studios/product-strategy/stages/prioritization/discovery/PRIORITY-MATRIX.md

@@ -0,0 +1,29 @@
+---
+name: priority-matrix
+location: .haiku/intents/{intent-slug}/knowledge/PRIORITY-MATRIX.md
+scope: intent
+format: text
+required: true
+---
+
+# Priority Matrix
+
+Scored and ranked opportunities with explicit trade-off documentation. This output feeds the roadmap stage as its primary input for sequencing decisions.
+
+## Content Guide
+
+Structure the matrix around defensible decision-making:
+
+- **Scoring framework** — dimensions used (impact, effort, strategic alignment, confidence, etc.) with weighting rationale
+- **Opportunity scores** — each opportunity scored per dimension with reasoning, not just numbers
+- **Ranked list** — final ordering with narrative explanation of why the top items win
+- **Trade-off documentation** — what is explicitly deprioritized and the reasoning behind it
+- **Stakeholder impact assessment** — anticipated reactions, feasibility concerns, and strategic alignment notes per top opportunity
+- **Confidence levels** — how certain each score is, and what would change it
+
+## Quality Signals
+
+- Scores have reasoning attached, not just numbers in a grid
+- Trade-offs are documented honestly — "we chose X over Y because..."
+- Confidence levels distinguish between data-backed scores and judgment calls
+- The matrix is actionable for roadmap creation, not just an analytical exercise

package/studios/product-strategy/stages/prioritization/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/product-strategy/stages/prioritization/hats/prioritizer.md

@@ -0,0 +1,8 @@
+**Focus:** Apply structured frameworks to score and rank opportunities. Make trade-offs explicit and defensible. The goal is a clear, reasoned ordering — not a mechanical score. Every ranking decision should have a "because" attached.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** treat framework scores as objective truth rather than structured judgment
+- The agent **MUST NOT** rank by a single dimension (impact only, effort only) without balancing factors
+- The agent **MUST NOT** fail to document the reasoning behind weights and scores
+- The agent **MUST NOT** hid low-confidence scores behind false precision
+- The agent **MUST NOT** avoid hard trade-offs by ranking everything as "high priority"

package/studios/product-strategy/stages/prioritization/hats/stakeholder-proxy.md

@@ -0,0 +1,8 @@
+**Focus:** Represent the perspectives of key stakeholders — business, engineering, sales, support — who are not in the room. Pressure-test prioritization against business constraints, resource realities, strategic commitments, and organizational politics. Surface the objections before stakeholders do.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** represent only one stakeholder group's perspective (e.g., only engineering feasibility)
+- The agent **MUST NOT** accept prioritization without challenging assumptions about effort or impact
+- The agent **MUST NOT** introduce stakeholder concerns as blockers instead of constraints to navigate
+- The agent **MUST NOT** project personal opinions as stakeholder positions without evidence
+- The agent **MUST NOT** ignore downstream effects on teams not directly involved in the decision

package/studios/product-strategy/stages/prioritization/hats/verifier.md

@@ -0,0 +1,26 @@
+**Focus:** Validate the per-unit design/synthesis artifact for the prioritization stage of product-strategy. Units here are prioritization decision — 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/product-strategy/stages/prioritization/outputs/PRIORITY-MATRIX.md

@@ -0,0 +1,25 @@
+---
+name: priority-matrix
+location: .haiku/intents/{intent-slug}/stages/prioritization/artifacts/
+scope: intent
+format: text
+required: true
+---
+
+# Priority Matrix
+
+Scored and ranked opportunities with trade-off documentation.
+
+## Expected Artifacts
+
+- **Scored opportunities** -- each opportunity scored on at least 3 dimensions with weighting rationale
+- **Rankings** -- top opportunities with impact estimates, effort estimates, and confidence levels
+- **Trade-off analysis** -- what is explicitly deprioritized and why
+- **Stakeholder validation** -- rankings pressure-tested against business constraints
+
+## Quality Signals
+
+- Each score includes rationale and confidence level
+- Rankings are pressure-tested against resource realities and strategic goals
+- Trade-offs are documented -- what we are choosing not to do and why
+- At least 3 scoring dimensions with explicit weighting

package/studios/product-strategy/stages/prioritization/phases/ELABORATION.md

@@ -0,0 +1,16 @@
+# Prioritization Stage — Elaboration
+
+## Criteria Guidance
+
+### Good criteria — concrete and verifiable
+
+- "Priority matrix scores each opportunity on at least 3 dimensions with explicit weighting rationale"
+- "Top 5 opportunities include impact estimates, effort estimates, and confidence levels"
+- "Trade-off analysis documents what is explicitly deprioritized and why"
+
+### Bad criteria — vague (no clear check)
+
+- "Priorities are set"
+- "Opportunities are ranked"
+- "Framework is applied"
+

package/studios/product-strategy/stages/prioritization/review-agents/rigor.md

@@ -0,0 +1,10 @@
+---
+interpretation: lens
+---
+**Mandate:** The agent **MUST** verify prioritization scores are defensible and free from bias.
+
+**Check:**
+- The agent **MUST** verify that scoring criteria are applied consistently across all opportunities
+- The agent **MUST** verify that impact and effort estimates are based on evidence, not gut feel
+- The agent **MUST** verify that stakeholder preferences do not override the framework without documented justification
+- The agent **MUST** verify that dependencies between opportunities are reflected in the priority order

package/studios/product-strategy/stages/roadmap/STAGE.md

@@ -0,0 +1,15 @@
+---
+name: roadmap
+description: Create the roadmap with sequencing, dependencies, and milestones
+hats: [roadmap-architect, capacity-planner, verifier]
+fix_hats: [roadmap-architect, feedback-assessor]
+review: ask
+elaboration: collaborative
+inputs:
+  - stage: prioritization
+    discovery: priority-matrix
+---
+
+# Roadmap
+
+Create the roadmap with sequencing, dependencies, and milestones.

package/studios/product-strategy/stages/roadmap/discovery/ROADMAP-DOC.md

@@ -0,0 +1,31 @@
+---
+name: roadmap-doc
+location: .haiku/intents/{intent-slug}/knowledge/ROADMAP-DOC.md
+scope: intent
+format: text
+required: true
+---
+
+# Roadmap Document
+
+The sequenced product roadmap with dependencies, milestones, and capacity validation. This output feeds the stakeholder-review stage as the primary artifact for presentation and alignment.
+
+## Content Guide
+
+Structure the roadmap as a strategic narrative:
+
+- **Strategic narrative** — the "why this order" story that connects phases to business outcomes
+- **Initiative sequencing** — ordered initiatives grouped into phases with dependency chains
+- **Dependency graph** — what blocks what, and what can run in parallel
+- **Milestones** — checkpoint definitions with measurable success criteria and constituent initiatives
+- **Capacity assessment** — resource mapping per phase, bottleneck identification, and feasibility verdict
+- **Assumptions and risks** — what must hold true for the roadmap to work, and contingency if it does not
+- **Flexibility points** — where the roadmap can adapt if priorities shift or assumptions break
+
+## Quality Signals
+
+- The roadmap tells a story, not just a sequence
+- Dependencies are explicit and the critical path is identified
+- Milestones have measurable criteria, not vague descriptions
+- Capacity constraints are reflected in the sequencing, not ignored
+- The document acknowledges what could change and how the roadmap would adapt

package/studios/product-strategy/stages/roadmap/hats/capacity-planner.md

@@ -0,0 +1,8 @@
+**Focus:** Validate the roadmap against resource constraints — team capacity, skills, infrastructure, and budget. Identify bottlenecks before they become surprises. Ensure the roadmap is ambitious but feasible, not aspirational fiction.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** rubber-stamp the roadmap without genuinely modeling capacity constraints
+- The agent **MUST NOT** treat all team members as interchangeable resources
+- The agent **MUST NOT** ignore ongoing operational work that competes for the same resources
+- The agent **MUST NOT** plan to 100% capacity with no slack for unplanned work
+- The agent **MUST NOT** flag every constraint as a blocker instead of proposing mitigation options

package/studios/product-strategy/stages/roadmap/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/product-strategy/stages/roadmap/hats/roadmap-architect.md

@@ -0,0 +1,8 @@
+**Focus:** Sequence prioritized opportunities into a coherent roadmap with dependency chains, milestones, and a strategic narrative. The roadmap should tell a story — why this order, what each phase unlocks, and how the pieces build on each other.
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** treat the roadmap as a flat list of features with arbitrary dates
+- The agent **MUST NOT** ignore dependencies between initiatives that constrain sequencing
+- The agent **MUST NOT** create milestones without measurable success criteria
+- The agent **MUST NOT** overpack phases without accounting for the unexpected
+- The agent **MUST NOT** build a roadmap that only works if every assumption holds — no flexibility for change

package/studios/product-strategy/stages/roadmap/hats/verifier.md

@@ -0,0 +1,26 @@
+**Focus:** Validate the per-unit design/synthesis artifact for the roadmap stage of product-strategy. Units here are roadmap entry — 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/product-strategy/stages/roadmap/outputs/ROADMAP-DOC.md

@@ -0,0 +1,25 @@
+---
+name: roadmap-doc
+location: .haiku/intents/{intent-slug}/stages/roadmap/artifacts/
+scope: intent
+format: text
+required: true
+---
+
+# Roadmap Document
+
+Sequenced initiatives with dependency graph, milestones, and capacity validation.
+
+## Expected Artifacts
+
+- **Initiative sequence** -- ordered initiatives with explicit dependency chains
+- **Dependency graph** -- visual representation of initiative dependencies
+- **Milestones** -- measurable success criteria for each milestone
+- **Capacity validation** -- resource constraints mapped to roadmap phases with bottlenecks flagged
+
+## Quality Signals
+
+- Each milestone has measurable success criteria
+- Dependency chains are explicit between initiatives
+- Capacity is validated against resource constraints
+- Roadmap tells a coherent narrative of why this sequence

package/studios/product-strategy/stages/roadmap/phases/ELABORATION.md

@@ -0,0 +1,16 @@
+# Roadmap Stage — Elaboration
+
+## Criteria Guidance
+
+### Good criteria — concrete and verifiable
+
+- "Roadmap sequences initiatives with explicit dependency chains between items"
+- "Each milestone has measurable success criteria and a list of constituent initiatives"
+- "Capacity plan identifies resource constraints and maps them to roadmap phases"
+
+### Bad criteria — vague (no clear check)
+
+- "Roadmap is created"
+- "Milestones are defined"
+- "Dependencies are noted"
+

package/studios/product-strategy/stages/roadmap/review-agents/feasibility.md

@@ -0,0 +1,10 @@
+---
+interpretation: lens
+---
+**Mandate:** The agent **MUST** verify the roadmap is achievable given capacity and dependencies.
+
+**Check:**
+- The agent **MUST** verify that sequencing respects technical dependencies
+- The agent **MUST** verify that capacity assumptions are realistic (not 100% utilization)
+- The agent **MUST** verify that external dependencies (APIs, partnerships, regulatory) have realistic lead times
+- The agent **MUST** verify that milestones have clear, verifiable completion criteria

package/studios/product-strategy/stages/stakeholder-review/STAGE.md

@@ -0,0 +1,15 @@
+---
+name: stakeholder-review
+description: Present to stakeholders, gather feedback, gain alignment
+hats: [presenter, feedback-synthesizer, verifier]
+fix_hats: [presenter, feedback-assessor]
+review: external
+elaboration: collaborative
+inputs:
+  - stage: roadmap
+    discovery: roadmap-doc
+---
+
+# Stakeholder Review
+
+Present to stakeholders, gather feedback, gain alignment.

package/studios/product-strategy/stages/stakeholder-review/discovery/ALIGNMENT-DOC.md

@@ -0,0 +1,31 @@
+---
+name: alignment-doc
+location: .haiku/intents/{intent-slug}/knowledge/ALIGNMENT-DOC.md
+scope: intent
+format: text
+required: true
+---
+
+# Alignment Document
+
+The final record of stakeholder review — decisions made, feedback incorporated, and ownership assigned. This output is the terminal artifact of the product-strategy lifecycle.
+
+## Content Guide
+
+Structure the document around decisions and accountability:
+
+- **Executive summary** — what was presented, what was decided, what changed
+- **Decision records** — each strategic decision with status (approved/contested/deferred), decision-maker, and rationale
+- **Feedback themes** — categorized stakeholder input mapped to roadmap decision points
+- **Roadmap adjustments** — changes made to the roadmap based on feedback, with before/after
+- **Contested items** — points of disagreement with escalation path or resolution plan
+- **Follow-up ownership** — who owns what, with clear next actions
+- **Open questions** — items that need further investigation before commitment
+
+## Quality Signals
+
+- Every decision has a named owner and clear status
+- Feedback is categorized by impact level, not just listed
+- Roadmap adjustments trace back to specific feedback, not vague "stakeholder input"
+- Contested items have explicit resolution paths, not hand-waves
+- The document is a reliable record — someone absent from the meeting can reconstruct what happened

package/studios/product-strategy/stages/stakeholder-review/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