GSD
Official@gsd-build
Offers modular engineering primitives for web quality assurance, frontend interface optimization, and systematic codebase maintenance.
Agent Skills by GSD
Showing 39 vetted skills indexed across 2 GitHub repositories.
btw
Answers quick side questions using only existing conversation context without tool calls.
create-mcp-server
Build and evaluate Model Context Protocol servers that expose APIs as LLM tools.
handoff
Writes a continue.md handoff file capturing current task state for cross-session resumption.
verify-before-complete
Enforces fresh verification output before any task completion claim.
api-design
Designs and reviews HTTP, REST, and GraphQL APIs with versioning, pagination, and error shapes.
observability
Adds structured logs, health endpoints, and persisted failure state to code for agent-first debugging.
forensics
Diagnose failed GSD auto-mode runs by tracing runtime artifacts to root causes.
decompose-into-slices
Breaks milestone plans into dependency-ordered vertical slices written to a roadmap file.
design-an-interface
Generates three divergent interface designs for a module and recommends one.
write-docs
Writes proposals, specs, ADRs, and README documentation through a three-stage reader-testing workflow.
spike-wrap-up
Convert completed research spike findings into reusable project-local agent skills.
write-milestone-brief
Synthesizes planning discussions into structured milestone brief documents following a fixed template.
dependency-upgrade
Plans and executes batched dependency upgrades with per-batch verification and isolated major-version commits.
tdd
Implements features through red-green-refactor cycles on vertical slices with behavior-pinning tests.
grill-me
Interviews users one question at a time to resolve every decision branch in a plan.
security-review
Performs STRIDE-based threat modeling and code-level security audits with severity-ranked findings.
gh
Install, configure, and operate the GitHub CLI for issues, PRs, milestones, and Projects V2.
gsd-browser
Control Chrome/Chromium browsers for web testing and data extraction.
gsd-orchestrator
Orchestrate GSD as a headless subprocess for autonomous software construction.
gsd-headless
Orchestrate GSD projects from a headless CLI with RPC child processes.
agent-browser
Automate browser actions and data collection via a CLI.
lint
Auto-detect and run the project's linter and formatter with optional fixes.
userinterface-wiki
Identify UI/UX best-practice gaps in web components and output file:line findings.
test
Detect test frameworks and generate or execute tests across Node.js, Python, and Go projects.
Frequently Asked Questions About GSD
FAQPage SchemaWhat specific tasks can GSD perform for web developers?▼
GSD enables automated web quality audits, including performance benchmarking for Core Web Vitals, accessibility scanning against WCAG 2.1 standards, and React component optimization. It also facilitates systematic code reviews, debugging hypothesis testing, and the generation of standardized interface components.
Which personas benefit most from these capabilities?▼
Frontend engineers, quality assurance specialists, and full-stack developers benefit from these capabilities. The system is designed for technical teams needing to enforce design-for-feel guidelines, maintain high-performance web interfaces, and ensure consistent code quality across complex repositories.
What are the prerequisites for running GSD components?▼
GSD components require a Node.js runtime environment and access to Chromium-based browsers for data extraction and interface testing. Users must define project structures using YAML frontmatter and XML body formats to integrate with the GSD engine and its associated headless subprocesses.