Orchestrate cross-agent task execution with TDD gating and telemetry collection.
npx skills add https://github.com/JeremyKalmus/parade --skill run-tasks
Please help me install this Agent Skill. Skill: run-tasks Source: https://github.com/JeremyKalmus/parade/tree/main/.claude/skills/run-tasks Command: npx skills add https://github.com/JeremyKalmus/parade --skill run-tasks
The run-tasks skill orchestrates cross-agent task execution within Claude Code projects by batching ready tasks, spawning specialized agents, and collecting telemetry to ensure reliable progress and completion.
Invoke the run-tasks skill to begin orchestrating an epic workflow in Claude Code. For example: /run-tasks <epic-id> to start processing ready tasks.
High-intent search queries and answers about installing and using this skill.
Cross-agent task orchestration batches ready tasks into sub-batches and runs them in parallel using git worktrees for isolation, spawning specialized agents to process tasks concurrently while collecting telemetry for retro analyses.
TDD gating enforces test-first development by gating implementation tasks behind tests, running RED/GREEN/DEBUG cycles to verify code, and using stop-hook enforcement to ensure task completion before proceeding.
Yes, git worktrees provide isolation for parallel batch execution, allowing agent-specific workers to operate independently on ready tasks without conflicts during cross-agent task orchestration.
Telemetry-driven retrospectives capture per-agent telemetry immediately after task completion, logging progress data that feeds retro analyses to improve future epic workflows and task batching strategies.
Beads metadata provides the task structure that run-tasks reads to drive automated execution, verification, and progress logging, determining which tasks are ready for batching and agent-specific worker spawning.
Stop-hook enforcement ensures task completion within the orchestration cycle, but it requires tasks to be properly gated by TDD verification and beads metadata to function effectively across parallel agent batches.
stevengonsalvezAutomate tmux-persistent multi-agent swarm creation from a Beads epic.
Wilder1222Evaluates sprint outcomes and generates improvement reports in docs/superomni/improvements/ for software teams.
DethonRun full verification commands and report fresh outputs before claiming completion.
jankneumannConvert OpenSpec proposals into Beads tasks and isolated Git worktrees.
