nelsonwerd
Community@nelsonwerd
nelsonwerd publishes composable skills for autonomous idea-to-ship pipelines: multi-agent codebase audits, iterative build verification, prompt decomposition, and concept validation.
Agent Skills by nelsonwerd
Showing 6 vetted skills indexed across 1 GitHub repositories.
nelsonwerddeep-dive
Orchestrates parallel specialist agents to audit codebases, strategies, designs, and research questions.
nelsonwerdaudit-and-fix
Audit an existing codebase and autonomously implement triaged fixes as verified local commits.
nelsonwerdbuild-loop
Iterates build, render, exercise, and critique cycles until acceptance criteria pass or a stop-condition fires.
nelsonwerdideate
Guides fuzzy ideas through a pressure-test funnel into a locked CONCEPT_BRIEF.md roadmap.
nelsonwerdprompt-pack
Decomposes large changes into sequenced, self-contained prompts executed across fresh chat sessions.
nelsonwerdautopilot
Orchestrates the idea-to-ship pipeline autonomously by composing ideate, deep-dive, prompt-pack, and build-loop skills.
Frequently Asked Questions About nelsonwerd
FAQPage SchemaWhat tasks can I accomplish using nelsonwerd's skills?▼
You can run rigorous multi-agent codebase audits with red-team review, autonomously implement audit findings as local commits, iterate builds until acceptance criteria pass, pressure-test product ideas into a locked CONCEPT_BRIEF.md, and decompose large jobs into sequenced prompt packs.
Who are nelsonwerd's skills designed for?▼
They target developers and technical founders who want structured, autonomous pipelines: engineers auditing unfamiliar repositories, builders driving drafts to near-finish-line quality, and solo founders validating whether an idea is worth building before writing code.
How do the skills run in practice?▼
Each skill is invoked by natural-language triggers like 'deep dive', 'audit and fix', or 'run autopilot'. They compose over existing capabilities such as bash build/test, headless screenshots with vision, Playwright, and axe/Lighthouse rather than reimplementing them.
What are the honest limitations of these skills?▼
Receipts record what ran but never prove correctness; build-loop self-grades taste with roughly an 80% craft ceiling; autopilot produces a near-finish-line first draft, not a market-validated product; and audit-and-fix makes local commits only, never pushing, merging, or publishing.
What prerequisites do nelsonwerd's skills require?▼
The skills are composition-first and depend on existing runtime capabilities: bash for build and test execution, headless screenshot plus vision tooling, Playwright, and axe/Lighthouse for UI verification. The autopilot skill additionally requires a real, grounded niche rather than invented demand.