Agent Skills by josephneumann
Showing 17 vetted skills indexed across 1 GitHub repositories.
josephneumannmulti-review
Orchestrate multi-agent code reviews across security, performance, and architecture.
josephneumannreconcile-summary
Reconciles worker session summaries with task board entries and updates statuses and descriptions based on divergences.
josephneumannproduct-review
Review product scope through EXPAND, HOLD, REDUCE, and DESIGN modes.
josephneumannorient
Analyze project state, documentation, and task queues to identify parallelizable work streams.
josephneumannverify
Enforce a five-step verification process for claims in agentic workflows.
josephneumannauto-run
Automate a dispatch-reconcile loop for batch task processing with agent teams.
josephneumanndispatch
Assign ready tasks to parallel agent teammates with risk assessment and model selection.
josephneumannspec
Transform feature descriptions into structured plans with decomposed tasks and acceptance criteria.
josephneumannretro
Analyze Git commit history to generate engineering retrospective reports.
josephneumannstart-task
Claim a task, gather project context, and define acceptance criteria.
josephneumannsummarize-session
Generate a read-only session summary from git context, tests, and task dependencies.
josephneumannwriting-skills
Create and refine SKILL.md files for AI agents.
josephneumannclaudemd-audit
Audit CLAUDE.md files for bloat, staleness, misplacement, and architectural issues.
josephneumannhumanizer
Rewrites AI-generated text to sound more natural and human.
josephneumannmilestone-review
Review and fix Git branch issues iteratively with parallel AI reviewers.
josephneumannfinish-task
Verify implementation, run quality gates, update documentation, and close a beads task.
josephneumanndebug
Guide systematic debugging through root cause analysis and hypothesis testing.