Harm
Community@HarmAalbers · Rotterdam
Harm maintains a Claude Code skill collection covering requirements frameworks, test-driven development, systematic debugging, refactoring orchestration, and Langfuse observability.
Agent Skills by Harm
Showing 22 vetted skills indexed across 1 GitHub repositories.
HarmAalberssession-learning
Analyzes Claude Code session metrics to recommend memory, skill, and command updates.
HarmAalbersusing-git-worktrees
Creates isolated git worktrees with directory selection, gitignore verification, and baseline test validation.
HarmAalbersrequirements-framework-status
Reports live version, component counts, and gating state of the Requirements Framework from repository commands.
HarmAalberstest-driven-development
Enforces red-green-refactor test-first workflow for writing features and bugfixes.
HarmAalberssystematic-debugging
Diagnose bugs through a four-phase root cause investigation process before proposing fixes.
HarmAalbersrefactor-orchestration
Plans and orchestrates multi-layer top-down refactors via chunked executor subagents.
HarmAalbersdispatching-parallel-agents
Dispatch concurrent agents to investigate independent failures across separate problem domains.
HarmAalbersexecuting-plans
Executes written implementation plans in batches with review checkpoints between batches.
HarmAalbersfinishing-a-development-branch
Guides branch completion by verifying tests and executing merge, PR, keep, or discard options.
HarmAalbersrequirements-framework-usage
Configure, customize, and troubleshoot the Claude Code Requirements Framework and its req CLI.
HarmAalbersrequirements-framework-development
Guides development, testing, and maintenance of the Claude Code Requirements Framework plugin.
HarmAalbersbrainstorming
Guides structured design interviews that turn feature ideas into approved design documents before implementation.
HarmAalbersworkflow-index
Maps the current requirements-framework phase to the next workflow command to run.
HarmAalbersusing-requirements-framework
Enforces skill discovery and invocation practices at conversation start within the requirements framework.
HarmAalberswriting-plans
Writes detailed implementation plans with bite-sized TDD tasks before any code is touched.
HarmAalbersrequirements-framework-builder
Extends the Claude Code Requirements Framework with custom strategies, calculators, and requirement types.
HarmAalbersrequesting-code-review
Dispatches code review agents to verify changes against requirements before merging.
HarmAalbersreceiving-code-review
Guides verification and implementation of code review feedback with technical reasoning.
HarmAalberswriting-skills
Create and test Claude Code skills using a TDD-based RED-GREEN-REFACTOR methodology.
HarmAalbersverification-before-completion
Enforces running verification commands before claiming work is complete or passing.
HarmAalberssubagent-driven-development
Executes implementation plans by dispatching fresh subagents per task with two-stage review.
HarmAalberslangfuse
Query and manage a self-hosted Langfuse instance via CLI and documentation.
Frequently Asked Questions About Harm
FAQPage SchemaWhat tasks can I accomplish with Harm's Claude Code skills?▼
The skills cover the full development lifecycle: brainstorming features, writing implementation plans, test-driven development, systematic debugging, orchestrating large refactors, dispatching parallel subagents, requesting and receiving code reviews, verifying completion, and managing a requirements framework with custom strategies and calculators.
Who are these skills designed for?▼
They target developers using Claude Code who want disciplined engineering practices: TDD before implementation, verification before claiming completion, git worktree isolation, and structured code review. They also serve maintainers of the claude-requirements-framework plugin and teams running self-hosted Langfuse observability.
How do the requirements framework skills work in practice?▼
The workflow-index skill maps your current phase to the next req command. The usage skill covers configuring checklists, scopes, and troubleshooting hooks, while the builder and development skills handle adding requirement types, custom strategies, calculators, and rebuilding the hooks-to-bundle pipeline with live-reload testing.
What does the Langfuse skill provide?▼
It enables querying and modifying a self-hosted Langfuse instance through langfuse-cli: listing and getting traces, prompts, scores, and sessions, inspecting the API schema, and fetching Langfuse documentation. It is tailored to the claude-requirements-framework's V3 observability stack including observability.py, prompts.py, and eval.py.
What are the prerequisites for using these skills?▼
Skills run inside Claude Code sessions and trigger on natural-language phrases like 'orchestrate refactor' or '/req'. The refactor orchestration skill uses Read, Write, Edit, Bash, Grep, Glob, WebFetch, and Context7 doc queries; the Langfuse skill requires a self-hosted Langfuse deployment plus npx or bunx langfuse-cli access.