TaskFlow
SkillProductivityUse when work should span one or more detached tasks but still behave like one job with one owner session, one return context, and durable flow state.
Available today. Use it from your connected AI after setup.
No other account needed.
Connect ahel once, and every AI you use reads what you have installed.
Then ask your AI: use the TaskFlow skill
What this skill tells your AI
The instructions your AI receives, as published by qianleigood/crawclaw in skills-optional/taskflow/SKILL.md and read by ahel’s review.
Use TaskFlow when a job needs to outlive one prompt or one detached run, but still needs one owner session and one place to inspect or resume the work.
Use it for
- multi-step background work with one owner
- jobs that wait on detached ACP or subagent tasks
- durable flow state between steps
- child-task linkage with revision-checked mutations
Do not use it for
- branching business logic
- routing policy
- domain-specific orchestration decisions
Keep those in the caller.
Canonical entrypoints
api.runtime.tasks.flowsapi.runtime.tasks.flows.fromToolContext(ctx)api.runtime.tasks.flows.bindSession({ sessionKey, requesterOrigin })
Lifecycle
createManaged(...)runTask(...)setWaiting(...)when blocked on a person or external systemresume(...)finish(...)orfail(...)requestCancel(...)orcancel(...)
Rules
- Use managed TaskFlows when your code owns orchestration.
- Treat
stateJsonas the persisted state bag. - Every mutating method after creation is revision-checked; always carry forward the latest revision.
- Use
runTask(...)instead of creating detached child tasks manually when you want parent orchestration. - Store only the minimum state needed to resume.
Signals
- GitHub stars
- 30
- Forks
- 1
- Last commit
- Sep 2026
Advanced
- Catalog kind
- skill
- Gateway key
taskflow-qianleigood- Source
- github.com/qianleigood/crawclaw