Resolve Task Parallel
Productivityresolve_task_parallelResolve Task Parallel is an Agent Skill for productivity published in the majiayu000/claude-skill-registry repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.
Resolve all pending CLI tasks using parallel processing
How to use this skill
Bring this guide into your coding agent with a prompt tailored to the tool you use.
- Open your project in Codex.
- Copy and paste the prompt below.
- Review files and risks before approving any action.
Inspect the Agent Skill at https://github.com/majiayu000/claude-skill-registry/blob/HEAD/skills/orchestration/resolve-task-parallel/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/resolve-task-parallel, but wait for my approval before writing files or running scripts.
Copying this prompt does not install or run the skill. Review third-party files before use.
Arguments
[optional: specific task ID or pattern]
Resolve tasks from the configured task system in parallel.
Workflow
1. Analyze
Use tasks-router to select the task system.
- If tasks-file: get unresolved tasks from
tasks/*.md - If tasks-beads: use
br list --status=open --jsonorbr ready --json
If any task recommends deleting, removing, or gitignoring files in docs/plans/ or docs/solutions/, skip it and mark it as wont_fix. These are workflow pipeline artifacts that are intentional and permanent.
2. Plan
Create a TodoWrite list (Claude) / update_plan list (Codex) of all unresolved tasks grouped by type. Identify dependencies and order the prerequisites first (for example, rename changes before usages). Output a mermaid flow diagram that shows what can run in parallel vs what must run sequentially.
3. Claim (Before Parallel Work)
- If tasks-beads: claim each task before starting:
br update <id> --claim --jsonThis setsstatus=in_progressandassignee=<actor>atomically.
4. Implement (PARALLEL)
Spawn one subagent per unresolved task item, in parallel. Each subagent should receive the specific task content and the expected outcome.
Example with 3 task items:
- Subagent for task1
- Subagent for task2
- Subagent for task3
Always run one subagent per unresolved task item, in parallel.
5. Commit & Resolve
- Commit changes
- If tasks-file: update the task file and mark it resolved
- If tasks-beads: close the issue with
br close <id> --reason="Completed" --json - Push to remote (when appropriate)