mattgierhartmattgierhartCommunityยท31 Agent Skills Included

PRD-driven-context-engineering

Gated PRD lifecycle with persistent product memory

Guides products from problem framing through launch using a gated v0.1-to-v1.0 PRD lifecycle with 47 stage-specific playbooks. Stores every decision as a traceable ID in markdown Source-of-Truth files, ending context loss between sessions and scattered documentation. Scores repo readiness automatically, blocks premature stage advancement, and coordinates four role agents with persistent memory.
npx skills add mattgierhart/PRD-driven-context-engineering --all -g -y
Available:

Tells the agent to load README, PRD, and Source-of-Truth files as its authority, treat the ID graph as the only truth, and route work through the four lifecycle role agents (strategy, design, build, launch).

All Skills in This Repository (31)

Pure Emerald Level Indicators
๐Ÿ“ฆ In Repo
mattgierhartmattgierhart

prd-v04-user-journey-mapping

Map triggers to value moments and generate UJ- entries for PRD v0.4.

Community
Advanced
๐Ÿ“ฆ In Repo
mattgierhartmattgierhart

prd-v09-feedback-loop-setup

Define feedback channels and triage rules for post-launch product improvement workflows.

Community
Advanced
๐Ÿ“ฆ In Repo
mattgierhartmattgierhart

prd-v07-test-planning

Plan TEST- entries with Given-When-Then templates linked to EPIC-, API-, BR-, and UJ- identifiers.

Community
Advanced
๐Ÿ“ฆ In Repo
mattgierhartmattgierhart

ghm-sot-builder

Scaffold standardized SoT files with YAML frontmatter and update protocols.

Community
Advanced
๐Ÿ“ฆ In Repo
mattgierhartmattgierhart

prd-v03-pricing-model

Select and validate pricing models for PRD v0.3 Commercial Model.

Community
Advanced
๐Ÿ“ฆ In Repo
mattgierhartmattgierhart

prd-v05-risk-discovery-interview

Identify and structure project risks through a guided PRD risk discovery interview.

Community
Advanced
๐Ÿ“ฆ In Repo
mattgierhartmattgierhart

prd-v04-persona-definition

Synthesize behavioral personas from CFD-, BR-, and FEA- evidence into traceable PER- entries.

Community
Intermediate
๐Ÿ“ฆ In Repo
mattgierhartmattgierhart

ghm-gate-check

Validate PRD gate criteria and generate consolidated pass/block reports.

Community
Advanced
๐Ÿ“ฆ In Repo
mattgierhartmattgierhart

prd-v07-epic-scoping

Transform PRD requirements into context-window sized EPICs with dependencies.

Community
Advanced
๐Ÿ“ฆ In Repo
mattgierhartmattgierhart

prd-v08-monitoring-setup

Define monitoring strategy, metrics, and alerting for PRD v0.8 deployments.

Community
Advanced
๐Ÿ“ฆ In Repo
mattgierhartmattgierhart

ghm-id-register

Validate and register new SoT IDs with format, uniqueness, and cross-reference checks.

Community
Advanced
๐Ÿ“ฆ In Repo
mattgierhartmattgierhart

prd-v08-release-planning

Generate DEP- entries and runbooks for structured deployment planning.

Community
Advanced

Frequently Asked Questions

FAQPage Schema
How to install PRD-driven-context-engineering?โ–ผ

Run `npx skills add mattgierhart/PRD-driven-context-engineering --all -g -y` in your terminal to install all skills globally. You can also fork the repo or run the included install script to seed the scaffold into an existing project.

What problem does PRD-driven context engineering solve?โ–ผ

It stops agents from forgetting prior decisions between sessions by storing every durable choice as a traceable ID in markdown Source-of-Truth files. It also prevents one-shot building by gating each PRD stage until its definition of done is met.

How does the gated PRD lifecycle work?โ–ผ

Work advances through stages v0.1 (problem framing) to v1.0 (market adoption), and a readiness scorer blocks advancement until required specs, tests, and cross-references exist. Each stage has dedicated skills like risk discovery, architecture design, and launch metrics.

Do I need coding experience to use this methodology?โ–ผ

No. Founders and product managers can drive the early stages (problem framing, personas, pricing, positioning) entirely through plain-English requests, while technical stages like architecture and testing activate when you reach them.

Which agents and frameworks are supported?โ–ผ

All skills follow the universal SKILL.md standard and run in Claude Code, Cursor, Codex, and OpenClaw. The four built-in role agents (Horizon, Studio, Devlab, Metro) coordinate strategy, design, build, and go-to-market work.

Related Repositories in Product & Management

View All in Product & Managementโ†’