Agent Skills by Nic
Showing 34 vetted skills indexed across 1 GitHub repositories.
cancel
Cancel active OMC modes with dependency-aware teardown and state cleanup.
autoresearch
Run evaluator-driven iterative improvement cycles until a bounded stop condition.
deep-dive
Produces a crop-parameters JSON file for a given image and input vector.
sciomc
Orchestrate parallel scientist-agent research stages into a verified synthesized report.
disk-cleanup
Recover lyric-lang dev environments from ENOSPC disk exhaustion by truncating files and reclaiming caches.
trace
Orchestrate competing hypotheses to produce evidence-driven causal explanations.
omc-reference
Retrieve OMC agent catalog, tools, team pipeline routing, and skills registry.
skill
Manage local Claude skill files by scanning, creating, editing, removing, searching, and syncing skill directories.
ask
Route prompts to Claude, Codex, or Gemini and persist responses as ask artifacts.
deep-interview
Clarify vague product ideas into testable specifications via Socratic interviews.
project-session-manager
Creates isolated git worktrees and optional tmux sessions for PR reviews, issues, and features.
self-improve
Run a research-to-plan-to-execute tournament loop to improve a target benchmark metric.
ultraqa
Automate iterative QA loops with failure diagnosis and fix application.
ultrawork
Coordinate parallel execution of independent tasks across parallel waves.
setup
Route setup, doctor, and MCP configuration requests by first argument token.
omc-setup
Configure oh-my-claudecode by installing CLAUDE.md and environment settings.
ralph
Iterate PRD stories until reviewer-verified acceptance criteria pass.
wiki
Maintain a persistent markdown knowledge base with keyword and tag queries.
deepinit
Generate hierarchical AGENTS.md documentation across repository directories.
omc-teams
Launch parallel claude, codex, or gemini CLI workers in tmux panes.
omc-doctor
Diagnose oh-my-claudecode installation issues by validating plugin versions and config remnants.
learner
Extract reusable debugging Skills from conversations with quality gates and YAML frontmatter.
ai-slop-cleaner
Simplify bloated AI-generated code while preserving intended behavior.
team
Coordinate up to 20 agents on a shared task graph with staged planning, execution, verification, and fixes.