Agent Skills by hmwkr
Showing 62 vetted skills indexed across 2 GitHub repositories.
HMWKRcynefin
Classify ambiguous problems into five Cynefin framework domains.
HMWKRproject-bootstrapper
Automate full-stack web and mobile project bootstrapping from a single request.
HMWKRharness-loop
Automate multi-agent orchestration loops with quality gates and Playwright verification.
HMWKRproject-kickstart
Automate a 7-step project bootstrapping pipeline and deliver artifacts as a ZIP archive.
HMWKRpipeline-orchestrator
Coordinate domain research, UX analysis, and project bootstrapping into a validated pipeline.
HMWKRthink-full
Enforce a 6-step sequential thinking workflow for strategic decisions.
HMWKRthink-deep
Analyze technical decisions through a structured 3-stage thinking workflow.
HMWKRultradetail-loop
Run infinite DOM enumeration and defect-fixing cycles with Playwright MCP.
HMWKRux-pattern-researcher
Collect domain-specific UI/UX patterns and competitor flows via structured web research.
HMWKRooda
Implement the OODA Loop adaptive cycle to adjust plans mid-task.
HMWKRthink-lite
Classify routine technical problems into Cynefin domains and run tailored thinking pipelines.
HMWKRthink-teams
Generate 3-6 domain-specific expert agents and execute a 6-stage parallel analysis pipeline.
HMWKRcantos-write
Generate standardized Cantos Architecture Decision Records and Design Decision Records.
HMWKRfirst-principles
Decompose decision premises into verified axioms through a structured 5-step workflow.
HMWKRdomain-researcher
Gather verified domain knowledge through automated web search and cross-validation.
HMWKRAgent Architect
Generate hierarchical agent team structures with CLAUDE.md and domain rules.
HMWKRinsight-sentinel
Scan git changes and commit history to detect unrecorded technical insights.
HMWKRlive-verify-loop
Run multi-round live browser verification with Playwright MCP until all layers pass.
HMWKRcontinuous-qa-loop
Orchestrate iterative multi-agent QA cycles to discover, fix, and verify codebase issues.
HMWKRplaywright-design-audit
Audit UI/UX and design quality across 450 checklist items with multi-agent teams.
HMWKRagent-teams-feature-dev
Automate full-stack feature development with parallel AI agent execution.
HMWKRunused-code-refactor-suggester
Scans source files for unused code and generates refactoring reports with risk summaries.
HMWKRdeep-analysis-mode
Analyze complex software engineering tasks using an 8-step workflow.
HMWKRinfra-audit
Audit Claude Code infrastructure and generate a 0-100 score with S/A/B/C/F grade.