skills.rest
ahuaracab
ahuaracabCOMMUNITY
Explore all skills by this author →
GitHub

project-discovery

Reverse-engineers an unknown codebase into PRD, SRS, glossary, and infrastructure context files.

Updated Aug 31, 2026
Search Tags:#srs#prd#project discovery#context generation#reverse engineering#codebase onboarding#qa framework
One-click install
npx skills add https://github.com/ahuaracab/test-genesis --skill project-discovery-ahuaracab
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: project-discovery
Source: https://github.com/ahuaracab/test-genesis/tree/main/.agents/skills/project-discovery
Command: npx skills add https://github.com/ahuaracab/test-genesis --skill project-discovery-ahuaracab

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Onboarding an unfamiliar repository for testing requires manually reconstructing its architecture, business model, and infrastructure. This Skill automates that reverse-engineering through four gated phases, producing structured context files that downstream QA skills consume. ## Core Features & Use Cases - Four-Phase Discovery Pipeline: Runs Constitution (business model, domain glossary), Architecture (PRD + SRS), Infrastructure (backend/frontend/CI-CD mapping), and Specification (backlog access) with completion gates and user confirmation between phases. - Session Resume & Subagent Orchestration: Persists plan and progress state so multi-hour discovery runs can resume after interruption, with per-phase subagent delegation for large codebases. - Fictional User Story Generation: When no issue tracker is available, synthesizes user stories from the feature map so brownfield repos still flow through the test pipeline. - Use Case: Point the Skill at a newly cloned Node.js/React monorepo; it detects the stack, documents the auth flow, database schema, and CI workflows, then hands off to context generation and framework adaptation. ## Quick Start Ask the assistant to onboard this repository by running project discovery against the target codebase path.

Frequently Asked Questions about project-discovery

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I onboard an unknown repository for test automation?▼

Run project-discovery against the local repo path. It executes four gated phases—Constitution, Architecture, Infrastructure, and Specification—producing business-model, PRD, SRS, and infrastructure files under .context/ that downstream testing skills consume.

What artifacts does project discovery generate from source code?▼

It generates business-model.md, domain-glossary.md, project-config.md, PRD documents (executive summary, personas, journeys), SRS documents (architecture, functional and non-functional specs), infrastructure docs, and a PBI backlog access recipe, each with a Discovery Gaps section.

Does project discovery work with monorepos?▼

Yes. It detects monorepos via pnpm-workspace.yaml, turbo.json, nx.json, or lerna.json, runs Phase 1 once at project level, and runs Phases 2-3 per package, merging findings into a single infrastructure directory with per-package subsections.

Can project discovery modify or write tests in the target repository?▼

No. Discovery is strictly read-only against the target repo; it only writes to the .context/ directory. Test authoring belongs to the test-automation skill, and framework wiring belongs to Adapt-framework.

What happens if the project has no issue tracker or empty backlog?▼

Phase 4.5 generates synthetic user stories from the business feature map, stored in fictional-hu-log.md with @fictional IDs. Downstream sprint-testing detects fictional_hu_mode and uses these records in place of real TMS tickets.

Why does project discovery refuse to regenerate business-data-map.md?▼

Business maps and the master test plan are owned by the separate project-context skill, which runs data, features, api, and test-plan modes in dependency order. Project-discovery only produces the foundational PRD/SRS/infrastructure artifacts and redirects map requests.

Related Skills

Explore More Software Engineering Skills→
shannonshellshannonshell
11

research-open-source

List, read, and search locally cloned repositories in a vendor directory.

Official
Intermediate
VirtoCommerceVirtoCommerce
2

qa-plan

Generate structured end-to-end test plans from scenario catalogs and feature requirements.

Official
Advanced
semicolon-devteamsemicolon-devteam

sync-interface

Generate TypeScript typings from the core-interface OpenAPI specification.

Official
Advanced
duylinhdang1998duylinhdang1998

business-analysis

Gather business requirements into structured SRS documents and Given/When/Then acceptance criteria.

Community
Advanced