bulk-run

Run Claude Code skills across multiple repositories in parallel with progress tracking.

2|Updated Jan 15, 2026
One-click install
npx skills add https://github.com/tsilva/claude-skills --skill bulk-run
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: bulk-run
Source: https://github.com/tsilva/claude-skills/tree/main/plugins/bulk-run/skills/bulk-run
Command: npx skills add https://github.com/tsilva/claude-skills --skill bulk-run

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill enables running a Claude Code skill across many repositories in parallel, dramatically speeding up large-scale automation tasks.

Core Features & Use Cases

  • Parallel execution: Dispatch the selected skill to multiple repositories concurrently with configurable batch sizes.
  • Progress tracking & resume: Persist progress to a local file and resume interrupted runs without repeating completed work.
  • Use Case: Run a README-generator across all repos in a monorepo to update documentation in one pass.

Quick Start

Use the bulk-run skill to execute another skill across all repos in a directory. Example: uv run bulk-run run --skill "readme-author" --repos-dir "/path/to/repos" --filter "mcp-*" --batch-size 4

Frequently Asked Questions about bulk-run

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run Claude Code skills across multiple repositories in parallel?

Parallel execution across repositories works by spawning individual Claude Code CLI processes for each repository concurrently. The script manages these processes in configurable batch sizes to optimize resource usage during large-scale codebase automation tasks.

Can I resume a bulk run if the Claude Code automation gets interrupted midway?

Yes, you can resume interrupted runs because the script saves progress to a local JSON file. This resume capability ensures that partially completed batch executions across repositories can continue without duplicating already finished work.

What is the best way to execute a skill on a filtered subset of my monorepo projects?

The best way to target a subset of repositories is by applying a filter pattern during the batch run. This ensures the Claude Code skill only executes in parallel across the directories matching your specified naming criteria.

Does bulk-run require any specific dependencies to batch process codebases?

Batch processing codebases with this tool requires Python and the Claude Code CLI installed in your environment. It operates without external package dependencies, using a local Python script to manage parallel execution and JSON state tracking.

How do I update documentation across all repositories in a monorepo at once?

You can update documentation across all repositories at once by dispatching a documentation skill in a parallel batch run. This applies the skill concurrently to each repository directory within the monorepo, completing the updates in a single pass.