Hunter Henrichsen
Community@hhenrichsen · Utah
Senior Software Engineer, @motion-canvas Maintainer, and Minecraft Plugin developer. All opinions/contributions my own unless otherwise stated.
Agent Skills by Hunter Henrichsen
Showing 27 vetted skills indexed across 1 GitHub repositories.
hhenrichsenshadcn--improve
Audits codebases and writes self-contained implementation plans for other agents to execute.
hhenrichsenponytail--ponytail
Enforces minimal, standard-library-first solutions for coding tasks with adjustable intensity levels.
hhenrichsenponytail--ponytail-audit
Audits an entire codebase for over-engineering and produces a ranked list of deletions and simplifications.
hhenrichsenponytail--ponytail-help
Displays a quick-reference card of ponytail modes, skills, and commands.
hhenrichsenponytail--ponytail-review
Reviews code diffs for over-engineering and lists deletable complexity with replacements.
hhenrichsenponytail--ponytail-gain
Displays a one-shot ASCII scoreboard of benchmark-measured code, cost, and speed reductions.
hhenrichsenponytail--ponytail-debt
Harvests ponytail: code comments into a grouped technical debt ledger report.
hhenrichsencaveman--caveman-help
Displays a quick-reference card of caveman modes, skills, and deactivation triggers.
hhenrichsencaveman--caveman-evidence-review
Reviews Caveman Cloud cost, trace, and savings evidence through read-only queries.
hhenrichsencaveman--investigate-first
Diagnoses ambiguous failures by ranking evidence-based hypotheses before editing code.
hhenrichsencaveman--caveman-explore
Locates relevant code in repositories and returns path:line citations.
hhenrichsencaveman--caveman-learn
Applies consent-gated fixes to ranked token sinks from a Caveman learn report.
hhenrichsencaveman--caveman-review
Generates one-line code review comments with location, severity, problem, and fix.
hhenrichsencaveman--caveman-compress
Compress markdown memory files into terse caveman format to reduce repeated input tokens.
hhenrichsencaveman--caveman-manage
Inspect Caveman Cloud experiment lifecycles and block unsafe lifecycle mutations.
hhenrichsencaveman--caveman-setup
Wire repository LLM callsites through the Caveman gateway for spend measurement.
hhenrichsencaveman--surgical-patch
Fix bugs at the narrowest responsible layer with regression proof.
hhenrichsencaveman--caveman-optimize
Evaluate Caveman report-only optimization observations with operator-chosen candidates and paired baseline evaluations.
hhenrichsencaveman--caveman-discover
Find and label LLM workflows in a repository for per-workflow spend tracking.
hhenrichsencaveman--safe-refactor
Restructure code while preserving behavior through verification-bracketed structural edits.
hhenrichsencaveman--verify-and-stop
Validates existing work against acceptance conditions without expanding scope.
hhenrichsencaveman--cavecrew
Guides delegation to compressed-output subagents for code location, editing, and review tasks.
hhenrichsencaveman--caveman-stats
Reports actual session token usage and estimated savings read from the JSONL session log.
hhenrichsencaveman--caveman
Compresses AI responses into terse caveman-style prose while preserving technical accuracy.
Frequently Asked Questions About Hunter Henrichsen
FAQPage SchemaWhat tasks can I accomplish with hhenrichsen's skills?▼
You can audit codebases for over-engineering and bloat, run compressed one-line code reviews, diagnose ambiguous failures with evidence-ranked hypotheses, apply surgical bug fixes, perform safe behavior-preserving refactors and reversible migrations, and measure or reduce LLM token spend through the Caveman Cloud gateway.
Who are these skills designed for?▼
Software engineers and senior developers who want minimal, standard-library-first solutions, teams fighting tech debt and speculative abstractions, and operators running LLM-assisted development who need per-workflow cost observability, token compression, and controlled experiment lifecycles.
How do the ponytail and caveman skills work in practice?▼
Invoke slash commands like /ponytail-review, /ponytail-audit, /caveman-compress, or /caveman-stats, or use natural triggers such as 'be lazy' or 'review this diff'. Most skills are one-shot read-only reports; edits like caveman-learn fixes require per-edit consent.
Are these skills open source and what do they cost?▼
The shadcn improve skill and ponytail minimalism skill are explicitly licensed under MIT, making them free to use and modify. Other caveman-family skills do not state a license in their frontmatter, though Caveman Cloud features involve a hosted gateway for spend measurement.
What prerequisites or dependencies do these skills require?▼
Most skills are self-contained prompt-driven modes needing no installation. caveman-explore declares Read, Glob, and Grep tools with the haiku model, and Caveman Cloud features (setup, evidence-review, manage, optimize) require wiring the repository through the Caveman Cloud gateway first.