Aaron Davis
Community@arndvs · ☀️San Diego
Aaron Davis maintains 40 skills for auditing, planning, and hardening codebases and agent-driven development sessions across TypeScript, Expo, and Sanity projects.
Agent Skills by Aaron Davis
Showing 40 vetted skills indexed across 1 GitHub repositories.
arndvsagent-navigability
Audit codebase structure for agent navigability and propose structural fixes or phased migrations.
arndvsresearch
Caches codebase exploration into persistent research documents before implementation begins.
arndvsagent-session-friction
Analyzes agent session transcripts to record measured codebase friction observations.
arndvsdo-work
Implements tasks through a plan, test, validate, commit, and preflight execution loop.
arndvsstress-test
Runs adversarial compliance tests against AI agent rule boundaries and produces scored audit reports.
arndvswrite-a-prd
Generates a structured Product Requirements Document from a rough feature idea through guided interviews.
arndvssystematic-debugging
Diagnose bugs through a four-phase root cause investigation process before proposing fixes.
arndvssanity-best-practices
Provides Sanity development best practices for schemas, GROQ queries, TypeGen, and framework integrations.
arndvshalbert-copy-editor
Edits sales copy, marketing emails, and landing pages using a four-stage copywriting method.
arndvssession-close
Audits session changes against quality gates before ending a coding session.
arndvscode-review
Reviews staged or recent git diffs to find edge cases, logic errors, and integration risks before merging.
arndvsfrontend-component-style
Enforces component file structure, naming conventions, and layer separation for React frontend code.
arndvsskill-scaffolder
Generates complete agent skill directories with orchestrators, state stores, and browser automation patterns.
arndvsnpm-security-audit
Audits npm, pnpm, and yarn projects for supply chain attack vectors before installation.
arndvsdocument
Write and audit technical documentation including READMEs, ADRs, changelogs, and Mermaid diagrams.
arndvsagent-assets
Audit committed agent commands, workflows, vendored docs, and tool configuration for gaps and staleness.
arndvscompliance-audit
Audits task diffs against active rules and skills to flag violations and log compliance.
arndvsimprove-architecture
Identifies shallow modules in a codebase and designs deeper module interfaces through parallel subagent analysis.
arndvsexplore
Investigates codebases by spawning parallel sub-agents for deep multi-file exploration.
arndvsperformance-audit
Diagnoses system and codebase performance issues including memory leaks, process bloat, and network latency.
arndvsagent-testability
Audit whether a codebase lets a coding agent verify its own changes and propose fixes.
arndvspress1-check
Audits Claude Code session logs to identify Bash commands that required manual approval.
arndvsdata-object-conventions
Enforces consistent structure, formatting, and placement for static data objects in TypeScript codebases.
arndvsmobile-dev
Diagnose and manage Expo React Native development sessions including Metro, emulators, and native rebuilds.
Frequently Asked Questions About Aaron Davis
FAQPage SchemaWhat tasks can I accomplish with Aaron Davis's skill collection?▼
You can audit codebases for bugs, architecture, logging, and performance; analyze agent session friction and navigability; write PRDs and break them into GitHub issues; run TDD red-green cycles; perform pre-PR reviews; and vet npm packages for security before installation.
Who is the target audience for these skills?▼
Software engineers and technical leads working in TypeScript/JavaScript, Expo/React Native, and Sanity codebases who use coding agents and want structured planning, review, debugging, and refactoring practices. Also suits teams measuring why agent runs are slow, costly, or unreliable.
How do the agent-focused auditing skills work in practice?▼
Skills like agent-navigability, agent-testability, and agent-session-friction inspect repo structure, test feedback loops, and completed session transcripts. They produce findings with measured cost or phased migration plans, helping you restructure repos so agents stop re-exploring code and can verify their own changes.
What prerequisites or environments do these skills assume?▼
Most skills operate on existing Git repositories with GitHub pull requests. Specific skills target npm/pnpm/yarn projects, Expo with Android emulator/QEMU, Sanity Studio schemas and GROQ, and Claude Code session transcripts. The TDD skill is backend-only and excludes frontend components.
Do these skills cover the full feature delivery lifecycle?▼
Yes. The manifest spans ideation (grill-me, write-a-prd), planning (architect, prd-to-issues), execution (do-work, tdd), verification (code-review, pr-preflight, compliance-audit), and cleanup (atomic-commits, session-close, plan-archive), forming an end-to-end delivery pipeline.