← ClaudeAtlas

agent-workflowslisted

Use when coordinating multi-agent work, parallel execution, task handoff, shared context, or clean session shutdown across multiple agents or terminals. Especially useful when the host supports subagents, hooks, or shared context. Common triggers: multi-agent, parallel work, agent coordination, task handoff, subagent delegation, version bump, bump version, release notes, changelog, tag release, release prep.
MarkBovee/nebu-skills · ★ 1 · AI & Automation · score 55
Install: claude install-skill MarkBovee/nebu-skills
# Nebu Agent Workflows ## Good fit Default to delegating auxiliary work. Spawn subagents freely; keep only what needs to persist in main context. - **any auxiliary work (default)** — grep, review, research, isolated edit - the work splits into independent parts - one branch is blocked on a slow command or external wait - a handoff between sessions is already happening - bounded release chore (version bump, changelog, release notes) ## Not a good fit - the steps are tightly coupled and need one shared thread of judgment → Gebruik in plaats daarvan `develop` staged delegation, die sequentiële dependency chains met per-stage validatie ondersteunt. - the next step depends directly on the exact output of the previous step - the reasoning or intermediate state is needed for the next step — losing it means re-deriving ## Context retention Default to delegate. Only keep in main when the reasoning must survive — structured output never needs to. | Keep in main | Delegate | |---|---| | Architecture / tradeoffs | grep, locate, map | | Reasoning chains (2+ steps) | Isolated edit, fixed spec | | Code that still changes | Bounded review | | Cross-cutting refactors | Research → summary | | Bug analysis needing full context | Mechanical rename, lint, format | ## Core lifecycle 1. Pick one current owner. 2. Split work only at clean boundaries. 3. Share the minimum context needed to avoid re-derivation. 4. Report blockers and findings early. 5. Keep the active owner driving tow