Agent Skills by fockus
Showing 32 vetted skills indexed across 1 GitHub repositories.
fockussadd:do-in-parallel
Dispatch parallel sub-agents across multiple targets with meta-judge verification.
fockuskaizen:kaizen
Apply kaizen principles to incrementally improve software quality and process design.
fockussadd:do-competitively
Orchestrate competitive multi-agent generation and multi-stage evaluation with a meta-judge.
fockussadd:do-and-judge
Coordinate parallel sub-agents and an automated judge to complete tasks with retries.
fockussdd:plan
Orchestrate a multi-agent workflow refining draft tasks into implementation-ready plans.
fockuskaizen:plan-do-check-act
Coordinate PDCA cycles for systematic experimentation and improvement.
fockuskaizen:cause-and-effect
Map root causes across six Fishbone categories to identify prioritized solutions.
fockusharness-sync
Synchronize Plans.md progress with implementation via drift detection and marker updates.
fockusharness-work
Automate end-to-end Harness v3 task execution from Plans.md workflows.
fockussadd:judge
Runs a two-phase evaluation producing an evidence-backed report with YAML specification.
fockuskaizen:analyse-problem
Generate a one-page A3 analysis with root cause and countermeasures.
fockusharness-setup
Orchestrate Harness v3 setup workflows for projects, tools, agents, and memory.
fockussadd:launch-sub-agent
Launch sub-agents with model selection and zero-shot CoT prompts.
fockusreflexion:memorize
Curate reflections and critique outputs into structured CLAUDE.md entries.
fockusharness-review
Automate multi-angle code, plan, and scope reviews for Harness v3 deployments.
fockusharness-release
Automate Harness v3 release workflows with version bumps, changelog updates, and GitHub Releases.
fockuskaizen:analyse
Select a Kaizen analysis technique and guide structured steps for a given target.
fockuskaizen:why
Document root causes through iterative Why analysis with corrective actions.
fockussadd:tree-of-thoughts
Coordinate parallel agents to explore, prune, and synthesize reasoning paths.
fockussadd:subagent-driven-development
Dispatch fresh subagents per task with code review between tasks.
fockussdd:brainstorm
Clarify rough ideas into concrete design concepts through guided brainstorming.
fockusbreezing
Coordinate multi-agent task execution across Lead, Worker, and Reviewer roles.
fockussadd:judge-with-debate
Coordinate three judge agents in iterative debates to produce consensus evaluations.
fockussadd:multi-agent-patterns
Distribute complex tasks across multiple agents with coordination patterns.