Dispatch team tasks serially to agents and persist parsed learnings.
npx skills add https://github.com/supertyrelle/pelley --skill sprint-supertyrelle
Please help me install this Agent Skill. Skill: sprint Source: https://github.com/supertyrelle/pelley/tree/main/skills/sprint Command: npx skills add https://github.com/supertyrelle/pelley --skill sprint-supertyrelle
This Skill removes the manual overhead of coordinating team execution by turning a plan or backlog into ordered agent assignments, tracked reflections, and persisted learnings.
Use the sprint skill to review the team state, assign the next ready task to the best-fit member, and persist any new learnings from the result.
High-intent search queries and answers about installing and using this skill.
You can automate task dispatch by loading a team manifest, reviewing learning history, and assigning ready tasks to the best-fit agent based on role fit and accumulated learnings, processing one task at a time to build sequential context.
Serial agent dispatch is a team coordination method that assigns one task at a time so each agent benefits from the context and findings of the previous task, ensuring execution order maintains continuity and captured reflections persist as learnings.
To persist learnings from agent task execution, parse the reflections generated after each assignment, validate the durable insights, and append the new learnings to each team member's individual memory file for future context.
Yes, task planning can integrate backlog triage and epic state by loading beads backlog state and prior epic findings alongside the team manifest, which improves assignment quality and determines optimal execution order.
Yes, a team manifest is required to assign tasks to agents because it provides the necessary team member roles, ownership details, and structure needed to match ready work to the correct agent and update their learning history.
A limitation of serial dispatch for backlog execution is that tasks are processed one at a time, which ensures sequential context but may reduce overall throughput compared to parallel agent assignment for independent backlog items.
supertyrelleAggregates read-only team status from git, learnings, manifests, and optional backlog state.
davidabeyerForces AI to extract learnings and suggest spec improvements before task completion.
phoenixvcSynchronize local backlog items with GitHub or Linear issue trackers.
tyevansOrchestrates a team of agents to execute tasks with learning loops and conflict resolution.
