Product Maestro turns a broad product-management discipline into a routed, testable package. Its value is structural: diagnosis, decisions, artifacts, and validation stay connected without claiming that a schema can prove market truth.
Public case study
Product Maestro
A production-grade product decision and delivery skill package for AI agents.
Active Python Schemas Validators Skill package Tests
Case study
Project breakdown
Product Maestro product flow
Repository-supported product flow for Product Maestro.
- 1 Request
- 2 Diagnose
- 3 Evaluate options
- 4 Define requirements
- 5 Quality review
Relationships
- Request Diagnose
- Diagnose Evaluate options
- Evaluate options Define requirements
- Define requirements Quality review
Request to Diagnose to Evaluate options to Define requirements to Quality review
Product Maestro architecture
Repository-supported architecture view for Product Maestro.
- Skill entrypoint
- References
- Schemas
- Validators
- Examples
Relationships
- Skill entrypoint References
- References Schemas
- Schemas Validators
- Validators Examples
Skill entrypoint to References to Schemas to Validators to Examples
Problem
Product Maestro provides a reusable operating structure for product diagnosis, discovery, prioritization, delivery, measurement, lifecycle, and review.
Overview
General AI guidance can jump from a request to an answer without preserving discovery, decision quality, delivery constraints, or governance.
Product reasoning
The package supports decision quality; it does not prove customer truth, product-market fit, adoption, commercial success, or the validity of evidence supplied to it.
Product decisions
Detailed material lives in routed references rather than one oversized prompt. Structured artifacts and deterministic validation protect the handoff between reasoning and execution.
Technical architecture
A compact skill entrypoint routes work into focused references, schemas, examples, validators, tests, and packaging tools.
Execution and delivery
The repository includes product templates, JSON schemas, positive and negative fixtures, validation commands, regression tests, architecture documentation, and an installable skill.zip artifact.
Current evidence
The public package can be installed as one skill and validated with its repository tooling and pytest suite. The repository also documents maintenance and release automation.
Learnings
A broad product discipline remains usable for an agent when routing is explicit and each major artifact has a verifiable contract.
Artifacts
Product evidence
documentation
Mandatory product decision sequence
The Skill entrypoint requires diagnosis, context, roles, evidence, opportunity, option and value evaluation, an explicit decision, requirements, measurement, governance, and quality review.
Substantive work is routed through decision gates before a PRD, roadmap, or delivery plan is treated as ready.
architecture
One entrypoint with routed references
The architecture separates agent instructions, progressively loaded references, schemas, templates, executable validators, examples, regression tests, and release automation.
Broad product knowledge is packaged behind one control plane without collapsing every domain into one oversized instruction file.
configuration
Versioned artifact contracts
The package defines supported product artifacts and requires selected JSON outputs to pass repository-owned schema and semantic validation.
Decision, evidence, requirements, measurement, release, governance, and quality handoffs have machine-checkable shapes.
Test and validation
test
Positive and negative regression coverage
The public tests validate examples, router output, readiness failure on critical quality ratings, requirement verification semantics, and the package audit.
Validation covers both accepted artifacts and failure conditions rather than schema parsing alone.
test
Pre-release integrity checklist
The checklist requires valid frontmatter, links, schemas and examples; negative tests; router checks; a zero-error audit; secret scanning; pinned automation; archive inspection; and version updates.
Release readiness includes package integrity, security hygiene, reproducibility, and content validation.
Release and distribution
release
Product Maestro v1.0.0 record
The changelog records the initial release with 24 deep references, 18 artifact schemas, validated examples, routing, semantic validation, audit tooling, tests, CI, and release packaging.
The public v1.0.0 scope is documented as a packaged, validated Skill rather than an unversioned prompt collection.
Current limitations
- The Skill supports product decision quality but cannot prove customer truth, product-market fit, adoption, commercial impact, or the validity of supplied evidence.
- It does not authorize production actions, accept controlled risks, or replace qualified legal, security, privacy, safety, compliance, financial, design, research, or engineering authority.
- No reviewable product screenshot is published upstream, so this Case Study does not fabricate one.
Overview
- Role
- Product architecture, scope definition, content system design, packaging, and validation
- Complexity
- High
- Detail level
- Case study
Responsibilities
- Problem definition
- Scope definition
- Product architecture
- Technical delivery
- Validation
Outcome
Product Maestro packages product decision and delivery guidance with schemas, examples, validators, tests, and an installable skill archive.