haiku-method 2.2.0

package/studios/hr/stages/offer/outputs/OFFER-PACKAGE.md

@@ -0,0 +1,26 @@
+---
+name: offer-package
+location: .haiku/intents/{intent-slug}/stages/offer/artifacts/
+scope: intent
+format: text
+required: true
+---
+
+# Offer Package
+
+Complete offer documentation including compensation analysis, offer letter, and negotiation parameters.
+
+## Content Guide
+
+- **Compensation analysis** -- market benchmarks, internal equity analysis, and positioning rationale
+- **Offer letter** -- complete terms including base, bonus, equity, start date, and any accommodations
+- **Negotiation parameters** -- approved flexibility ranges for key terms
+- **Contingency plan** -- fallback candidates and timeline if primary candidate declines
+- **Approval record** -- all required approvals documented with dates
+
+## Quality Signals
+
+- Compensation is benchmarked against both external market data and internal equity
+- Offer letter includes all material terms with no ambiguity
+- Negotiation parameters are pre-approved to enable responsive candidate engagement
+- Contingency plan is actionable with named fallback candidates

package/studios/hr/stages/offer/phases/ELABORATION.md

@@ -0,0 +1,16 @@
+# Offer Stage — Elaboration
+
+## Criteria Guidance
+
+### Good criteria — concrete and verifiable
+
+- "Compensation analysis benchmarks the candidate against internal equity bands and external market data with documented rationale for positioning"
+- "Offer letter includes all material terms: base, bonus, equity, start date, and any negotiated accommodations"
+- "Contingency plan documents fallback candidates and timeline if the primary candidate declines"
+
+### Bad criteria — vague (no clear check)
+
+- "Offer is sent"
+- "Compensation is fair"
+- "Candidate accepted"
+

package/studios/hr/stages/offer/review-agents/equity.md

@@ -0,0 +1,10 @@
+---
+interpretation: lens
+---
+**Mandate:** The agent **MUST** verify compensation package is equitable and properly benchmarked.
+
+**Check:**
+- The agent **MUST** verify that compensation is benchmarked against external market data and internal equity bands
+- The agent **MUST** verify that package positioning within the range has documented rationale
+- The agent **MUST** verify that no internal equity violations are created by the offer
+- The agent **MUST** verify that all approval workflows are complete before offer extension

package/studios/hr/stages/requisition/STAGE.md

@@ -0,0 +1,13 @@
+---
+name: requisition
+description: Define role requirements and create job description
+hats: [hiring-manager, recruiter, verifier]
+fix_hats: [hiring-manager, feedback-assessor]
+review: ask
+elaboration: collaborative
+inputs: []
+---
+
+# Requisition
+
+Define role requirements and create job description.

package/studios/hr/stages/requisition/discovery/JOB-SPEC.md

@@ -0,0 +1,29 @@
+---
+name: job-spec
+location: .haiku/intents/{intent-slug}/knowledge/JOB-SPEC.md
+scope: intent
+format: text
+required: true
+---
+
+# Job Spec
+
+Complete role specification including requirements, compensation benchmarks, and hiring context.
+
+## Content Guide
+
+Structure the job spec for both sourcing and evaluation:
+
+- **Business case** -- why this role exists and what team gap or need it addresses
+- **Role responsibilities** -- specific duties and expected outcomes
+- **Must-have qualifications** -- requirements that are genuinely necessary with rationale
+- **Nice-to-have qualifications** -- preferred qualities that would enhance the candidate
+- **Compensation benchmarks** -- market data from multiple sources with geographic adjustments
+- **Hiring context** -- timeline, urgency, team dynamics, and growth opportunity
+
+## Quality Signals
+
+- Must-have requirements are genuinely necessary, not aspirational wish lists
+- Compensation range is benchmarked against current market data
+- Role responsibilities focus on outcomes, not just activities
+- The spec provides enough detail for effective sourcing and screening

package/studios/hr/stages/requisition/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/hr/stages/requisition/hats/hiring-manager.md

@@ -0,0 +1,13 @@
+**Focus:** Define the business need for the role, articulate required competencies, and set expectations for the hiring process.
+
+**Responsibilities:**
+- Document the business case for the role with specific team gaps or growth needs
+- Define must-have versus nice-to-have qualifications with rationale
+- Set compensation expectations aligned with budget and market data
+- Establish the hiring timeline and urgency level
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** define the role as a clone of the last person in the seat without reassessing needs
+- The agent **MUST NOT** list requirements that are aspirational rather than actually needed for success
+- The agent **MUST NOT** set compensation ranges without market benchmarking
+- The agent **MUST** involv the team in defining what the role actually needs

package/studios/hr/stages/requisition/hats/recruiter.md

@@ -0,0 +1,13 @@
+**Focus:** Translate business requirements into an effective job description and validate that requirements are realistic for the target candidate market.
+
+**Responsibilities:**
+- Craft job descriptions that accurately represent the role and attract qualified candidates
+- Validate requirements against market availability and compensation benchmarks
+- Define sourcing strategy based on role requirements and market conditions
+- Identify potential challenges in the hiring process early
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** copy job descriptions from competitors without adapting to the actual role
+- The agent **MUST NOT** include requirements that systematically exclude qualified candidates
+- The agent **MUST NOT** write descriptions that focus on what the company wants without articulating what it offers
+- The agent **MUST** challeng unrealistic requirements from the hiring manager

package/studios/hr/stages/requisition/hats/verifier.md

@@ -0,0 +1,26 @@
+**Focus:** Validate the per-unit knowledge artifact for the requisition stage of hr. Units here are requisition artifact — 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/hr/stages/requisition/outputs/JOB-SPEC.md

@@ -0,0 +1,25 @@
+---
+name: job-spec
+location: .haiku/intents/{intent-slug}/stages/requisition/artifacts/
+scope: intent
+format: text
+required: true
+---
+
+# Job Specification
+
+Role definition with responsibilities, qualifications, compensation benchmarks, and hiring timeline.
+
+## Expected Artifacts
+
+- **Role description** -- responsibilities mapped to specific team gaps or business needs
+- **Qualifications** -- must-have vs nice-to-have with rationale for each requirement
+- **Compensation benchmarks** -- benchmarked against market data with geographic adjustments
+- **Hiring timeline** -- target dates for each stage of the hiring process
+
+## Quality Signals
+
+- Role addresses a validated business need with supporting evidence
+- Requirements are realistic for the target candidate market
+- Compensation is benchmarked against at least 3 market data sources
+- Must-have and nice-to-have qualifications are clearly distinguished

package/studios/hr/stages/requisition/phases/ELABORATION.md

@@ -0,0 +1,16 @@
+# Requisition Stage — Elaboration
+
+## Criteria Guidance
+
+### Good criteria — concrete and verifiable
+
+- "Job spec includes must-have vs nice-to-have qualifications with clear rationale for each requirement"
+- "Compensation range is benchmarked against at least 3 market data sources with geographic adjustments"
+- "Role requirements map to specific team gaps or business needs with supporting evidence"
+
+### Bad criteria — vague (no clear check)
+
+- "Job description is written"
+- "Requirements are defined"
+- "Role is approved"
+

package/studios/hr/stages/requisition/review-agents/market-fit.md

@@ -0,0 +1,10 @@
+---
+interpretation: lens
+---
+**Mandate:** The agent **MUST** verify role requirements are realistic for the target candidate market.
+
+**Check:**
+- The agent **MUST** verify that must-have requirements are genuinely necessary, not aspirational
+- The agent **MUST** verify that compensation range is benchmarked against current market data
+- The agent **MUST** verify that job description accurately represents the role without over-inflation
+- The agent **MUST** verify that requirements do not unnecessarily narrow the candidate pool

package/studios/hr/stages/screening/STAGE.md

@@ -0,0 +1,17 @@
+---
+name: screening
+description: Resume review and initial candidate qualification
+hats: [screener, assessor]
+fix_hats: [screener, feedback-assessor]
+review: auto
+elaboration: autonomous
+inputs:
+  - stage: sourcing
+    discovery: candidate-pipeline
+  - stage: requisition
+    discovery: job-spec
+---
+
+# Screening
+
+Resume review and initial candidate qualification.

package/studios/hr/stages/screening/discovery/SCREENING-REPORT.md

@@ -0,0 +1,28 @@
+---
+name: screening-report
+location: .haiku/intents/{intent-slug}/knowledge/SCREENING-REPORT.md
+scope: intent
+format: text
+required: true
+---
+
+# Screening Report
+
+Candidate evaluation results with consistent scoring and ranked shortlist.
+
+## Content Guide
+
+Structure the report for interview stage planning:
+
+- **Screening methodology** -- criteria used, scoring methodology, and consistency checks
+- **Candidate scores** -- each candidate scored against must-have and nice-to-have criteria
+- **Ranked shortlist** -- candidates ordered by composite fit score with rationale
+- **Disqualification summary** -- candidates not advancing with specific reasons
+- **Pool observations** -- patterns in the candidate pool that may inform sourcing adjustment
+
+## Quality Signals
+
+- All candidates are evaluated against identical criteria
+- Scoring methodology is transparent and consistently applied
+- Disqualification reasons trace to specific job spec requirements
+- The shortlist includes sufficient candidates for the interview stage

package/studios/hr/stages/screening/hats/assessor.md

@@ -0,0 +1,13 @@
+**Focus:** Validate screening consistency and produce a ranked candidate shortlist for the interview stage.
+
+**Responsibilities:**
+- Review screener decisions for consistency and calibration
+- Score candidates on a composite fit metric using weighted criteria
+- Produce a ranked shortlist with supporting rationale for each candidate
+- Identify patterns in the candidate pool that may require sourcing adjustment
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** override screener decisions without documented rationale
+- The agent **MUST NOT** rank candidates without a transparent, repeatable methodology
+- The agent **MUST NOT** advanc too many or too few candidates for the interview stage
+- The agent **MUST NOT** ignore candidate pool composition patterns that signal sourcing issues

package/studios/hr/stages/screening/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/hr/stages/screening/hats/screener.md

@@ -0,0 +1,13 @@
+**Focus:** Systematically evaluate candidates against job requirements using consistent criteria.
+
+**Responsibilities:**
+- Apply screening criteria consistently across all candidates
+- Document pass/fail decisions with specific evidence from resumes and profiles
+- Flag edge cases where candidates are borderline for assessor review
+- Maintain screening velocity to keep the hiring timeline on track
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** apply different standards to different candidates based on source or referral
+- The agent **MUST NOT** make screening decisions based on gut feel without reference to criteria
+- The agent **MUST NOT** reject candidates for missing nice-to-have items when must-haves are met
+- The agent **MUST** document the rationale for pass/fail decisions

package/studios/hr/stages/screening/outputs/SCREENING-REPORT.md

@@ -0,0 +1,25 @@
+---
+name: screening-report
+location: .haiku/intents/{intent-slug}/stages/screening/artifacts/
+scope: intent
+format: text
+required: true
+---
+
+# Screening Report
+
+Candidate evaluation results with scoring, ranking, and qualification decisions.
+
+## Expected Artifacts
+
+- **Candidate scores** -- each candidate scored against must-have criteria with pass/fail justification
+- **Ranking** -- candidates ranked by composite fit score with clear methodology
+- **Disqualification rationale** -- specific reasons traceable to job spec requirements
+- **Shortlist** -- candidates advancing to interview with supporting justification
+
+## Quality Signals
+
+- Every candidate is scored against must-have criteria with documented justification
+- Disqualification reasons are specific and traceable to job spec requirements
+- Ranking methodology is consistent and transparent
+- Shortlist size is appropriate for the interview capacity

package/studios/hr/stages/screening/phases/ELABORATION.md

@@ -0,0 +1,16 @@
+# Screening Stage — Elaboration
+
+## Criteria Guidance
+
+### Good criteria — concrete and verifiable
+
+- "Each candidate is scored against must-have criteria with pass/fail justification documented"
+- "Screening report ranks candidates by composite fit score with clear methodology"
+- "Disqualification reasons are specific and traceable to job spec requirements, not subjective impressions"
+
+### Bad criteria — vague (no clear check)
+
+- "Candidates are screened"
+- "Top candidates are identified"
+- "Resumes are reviewed"
+

package/studios/hr/stages/screening/review-agents/consistency.md

@@ -0,0 +1,10 @@
+---
+interpretation: lens
+---
+**Mandate:** The agent **MUST** verify screening decisions are consistent and traceable to job requirements.
+
+**Check:**
+- The agent **MUST** verify that all candidates are evaluated against the same criteria
+- The agent **MUST** verify that pass/fail decisions reference specific job spec requirements
+- The agent **MUST** verify that scoring methodology is applied consistently across the candidate pool
+- The agent **MUST** verify that edge cases are flagged for additional review rather than arbitrarily decided

package/studios/hr/stages/sourcing/STAGE.md

@@ -0,0 +1,15 @@
+---
+name: sourcing
+description: Identify candidate pools and conduct outreach
+hats: [sourcer, recruiter, verifier]
+fix_hats: [sourcer, feedback-assessor]
+review: auto
+elaboration: autonomous
+inputs:
+  - stage: requisition
+    discovery: job-spec
+---
+
+# Sourcing
+
+Identify candidate pools and conduct outreach.

package/studios/hr/stages/sourcing/discovery/CANDIDATE-PIPELINE.md

@@ -0,0 +1,28 @@
+---
+name: candidate-pipeline
+location: .haiku/intents/{intent-slug}/knowledge/CANDIDATE-PIPELINE.md
+scope: intent
+format: text
+required: true
+---
+
+# Candidate Pipeline
+
+Sourced candidate pool with initial fit assessments and channel effectiveness data.
+
+## Content Guide
+
+Structure the pipeline for efficient screening:
+
+- **Pipeline summary** -- total candidates by source channel with response rates
+- **Candidate profiles** -- for each candidate: source, relevant experience, initial fit assessment
+- **Channel effectiveness** -- which channels produced the most and best candidates
+- **Outreach results** -- response rates, engagement levels, and lessons learned
+- **Pipeline health** -- volume sufficiency assessment against hiring timeline
+
+## Quality Signals
+
+- Pipeline includes candidates from multiple distinct sourcing channels
+- Each candidate has an initial fit assessment against job spec requirements
+- Channel effectiveness is tracked with quantitative metrics
+- Pipeline volume is sufficient to support the screening stage's needs

package/studios/hr/stages/sourcing/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/hr/stages/sourcing/hats/recruiter.md

@@ -0,0 +1,13 @@
+**Focus:** Manage the sourcing process, ensure pipeline quality, and maintain candidate experience throughout outreach.
+
+**Responsibilities:**
+- Review sourced candidates for initial fit against job spec requirements
+- Manage candidate communication and ensure timely responses
+- Monitor pipeline health metrics (volume, quality, response rates)
+- Coordinate with the sourcer to adjust channel strategy based on results
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** let sourced candidates go uncontacted for extended periods
+- The agent **MUST** provide feedback to the sourcer on candidate quality from each channel
+- The agent **MUST NOT** treat sourcing as a one-time activity rather than an ongoing pipeline effort
+- The agent **MUST NOT** fail to track which channels produce the best conversion rates

package/studios/hr/stages/sourcing/hats/sourcer.md

@@ -0,0 +1,13 @@
+**Focus:** Identify and engage potential candidates through diverse sourcing channels, building a qualified pipeline.
+
+**Responsibilities:**
+- Research and identify candidate pools across multiple channels (networks, platforms, referrals)
+- Create personalized outreach tailored to each candidate's background
+- Track sourcing channel effectiveness to optimize effort allocation
+- Build pipeline volume sufficient to meet hiring timeline
+
+**Anti-patterns (RFC 2119):**
+- The agent **MUST NOT** rely on a single sourcing channel for the entire pipeline
+- The agent **MUST NOT** send generic outreach that does not reference the candidate's specific background
+- The agent **MUST NOT** source for volume without qualifying candidates against core requirements
+- The agent **MUST NOT** ignore passive candidates who are not actively job-seeking

package/studios/hr/stages/sourcing/hats/verifier.md

@@ -0,0 +1,26 @@
+**Focus:** Validate the per-unit operational artifact for the sourcing stage of hr. Units here are sourcing batch — 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/hr/stages/sourcing/outputs/CANDIDATE-PIPELINE.md

@@ -0,0 +1,25 @@
+---
+name: candidate-pipeline
+location: .haiku/intents/{intent-slug}/stages/sourcing/artifacts/
+scope: intent
+format: text
+required: true
+---
+
+# Candidate Pipeline
+
+Qualified prospect list from diverse sourcing channels with initial fit assessments.
+
+## Expected Artifacts
+
+- **Prospect list** -- qualified candidates from 3+ distinct sourcing channels
+- **Candidate profiles** -- source, relevant experience match, and initial fit assessment per candidate
+- **Channel effectiveness** -- metrics on which sourcing channels produced the best results
+- **Outreach records** -- personalized outreach sent with response tracking
+
+## Quality Signals
+
+- Pipeline includes at least 20 qualified prospects
+- Candidates come from 3+ distinct sourcing channels
+- Each profile documents source and initial fit against the job spec
+- Pipeline volume is sufficient to meet the hiring timeline

package/studios/hr/stages/sourcing/phases/ELABORATION.md

@@ -0,0 +1,16 @@
+# Sourcing Stage — Elaboration
+
+## Criteria Guidance
+
+### Good criteria — concrete and verifiable
+
+- "Candidate pipeline includes at least 20 qualified prospects from 3+ distinct sourcing channels"
+- "Each candidate profile documents source, relevant experience match, and initial fit assessment"
+- "Outreach messages are personalized to each candidate's background and the specific role value proposition"
+
+### Bad criteria — vague (no clear check)
+
+- "Candidates are found"
+- "Pipeline is full"
+- "Outreach is done"
+

package/studios/hr/stages/sourcing/review-agents/diversity.md

@@ -0,0 +1,10 @@
+---
+interpretation: lens
+---
+**Mandate:** The agent **MUST** verify candidate pipeline is sourced from diverse channels and represents a varied candidate pool.
+
+**Check:**
+- The agent **MUST** verify that pipeline includes candidates from multiple sourcing channels
+- The agent **MUST** verify that outreach is personalized and professional across all channels
+- The agent **MUST** verify that pipeline volume is sufficient to meet hiring timeline requirements
+- The agent **MUST** verify that channel effectiveness is tracked for future sourcing optimization

package/studios/hwdev/STUDIO.md

@@ -0,0 +1,25 @@
+---
+name: hardware-development
+slug: hwdev
+description: Lifecycle for hardware products — electronics, firmware, manufacturing
+stages: [inception, requirements, design, firmware, validation, manufacturing]
+category: engineering
+default_model: sonnet
+---
+
+# Hardware Development
+
+Lifecycle for hardware products. Unlike application development, hardware has
+physical constraints, safety regulations, one-shot manufacturing tooling, and
+a cost structure where late changes are orders of magnitude more expensive
+than early ones.
+
+Inception is market research (same shape as application development).
+Requirements captures functional, safety, and regulatory constraints upfront
+because they shape every downstream decision and cannot be retrofitted.
+
+The design stage uses [tscircuit](https://tscircuit.com) as the EDA platform —
+schematics and PCB layouts are authored as TypeScript/React (`.tsx`) circuit
+code, previewed live via `tsci dev`, and exported to Gerbers, pick-and-place,
+and BOM files for manufacturing. This keeps electronics design diff-able,
+reviewable in a pull request, and reproducible from source.

package/studios/hwdev/examples/indoor-air-sensor/intent.md

@@ -0,0 +1,18 @@
+---
+title: indoor-air-sensor
+studio: hwdev
+stages: [inception, requirements, design, firmware, validation, manufacturing]
+status: complete
+---
+
+# Indoor Air Quality Sensor
+
+Battery-powered indoor air-quality sensor. Measures temperature, humidity, and CO₂. 12-month battery life on 2×AA. BLE for host connectivity.
+
+## Goals
+
+- 12-month battery life at 15-minute reporting cadence
+- NDIR CO₂ sensor for measurement accuracy
+- BLE 5 for low-power host connectivity
+- FCC / CE compliance from day one
+- OTA firmware updates over BLE

package/studios/hwdev/examples/indoor-air-sensor/pre-intent.yml

@@ -0,0 +1,17 @@
+related:
+  - compliance
+  - data-pipeline
+  - software
+turns:
+  - role: user
+    text: "Design a battery-powered environmental sensor."
+  - role: agent
+    text: "Which measurements — temp/humidity, VOC, CO₂, or a custom mix?"
+  - role: user
+    text: "Temp, humidity, CO₂. Indoor air quality focus."
+  - role: agent
+    text: "Battery-life target and wireless — LoRa, BLE, or WiFi?"
+  - role: user
+    text: "12 months on 2×AA, BLE."
+  - role: agent
+    text: "Got it — I'll start the inception brief with those constraints."

package/studios/hwdev/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/hwdev/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