Agent Skills by Matt Niedelman
Showing 19 vetted skills indexed across 1 GitHub repositories.
mattniedelmanresearch
Automate structured research and generate markdown reports saved to Basic Memory.
mattniedelmanusing-git-worktrees
Create isolated git worktrees with new branches and verify clean baselines.
mattniedelmantest-driven-development
Guide teams through Red-Green-Refactor cycles for features, bug fixes, and refactors.
mattniedelmansystematic-debugging
Guide engineers through a four-phase root-cause debugging process.
mattniedelmanusing-superpowers
Check for applicable skills before responding to any user message.
mattniedelmandispatching-parallel-agents
Parallelize independent debugging investigations into structured summaries.
mattniedelmanexecuting-plans
Execute written plans in controlled batches with review checkpoints.
mattniedelmanfinishing-a-development-branch
Validate tests and present merge, PR, keep, or discard options for a development branch.
mattniedelmanspec-driven-development
Execute SPEC-1 workflows to create, implement, validate, and document feature specs.
mattniedelmancontinue-conversation
Resumes previous work by building context from the Basic Memory knowledge graph.
mattniedelmanbrainstorming
Convert vague ideas into structured design briefs with guided inquiry.
mattniedelmanknowledge-capture
Capture conversation insights into Basic Memory knowledge graph notes.
mattniedelmanwriting-plans
Convert high-level specs into structured implementation plans with tasks and file references.
mattniedelmanrequesting-code-review
Dispatch a reviewer subagent to evaluate recent code changes.
mattniedelmanreceiving-code-review
Clarify and verify code-review feedback before implementing changes with tests.
mattniedelmanknowledge-organize
Identify orphan notes and propose relations in large note collections.
mattniedelmanwriting-skills
Creates, edits, and verifies process-writing skills with YAML frontmatter and Markdown bodies.
mattniedelmanverification-before-completion
Require fresh command outputs as evidence before claiming task completion.
mattniedelmansubagent-driven-development
Dispatch fresh subagents for task implementation with two-stage spec and code-quality reviews.