skills.rest
tigredonorte
tigredonorteCOMMUNITY
Explore all skills by this author →
GitHub

run-agent

Runs Claude agents with safety hooks, turn limits, and restricted tools via CLI.

Updated Feb 25, 2026
Search Tags:#automation#cli#claude#agent#task-management#hook-protection#turn-limits
One-click install
npx skills add https://github.com/tigredonorte/claude-plugin-work --skill run-agent
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: run-agent
Source: https://github.com/tigredonorte/claude-plugin-work/tree/main/skills/run-agent
Command: npx skills add https://github.com/tigredonorte/claude-plugin-work --skill run-agent

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Run a Claude agent in a controlled, safe manner with built-in hook protection and strict turn limits to prevent runaway conversations.

Core Features & Use Cases

  • Deterministic CLI-based agent execution with optional background run
  • Enforces max turns, restricts allowed tools, and disables session persistence to ensure repeatable results
  • Supports multiple agent types (qa, code-review, explore, implement) for diverse tasks such as QA testing, code reviews, exploration, and implementation

Quick Start

Run the QA agent on a task by typing /run-agent qa <your-task-description>.

Frequently Asked Questions about run-agent

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

FAQPage Schema
How do I run a safe Claude agent with hook protection and turn limits?▼

To run a safe Claude agent, you can use the run-agent CLI to execute controlled tasks with safety hooks, enforcing max turns, restricted tools, and no session persistence to prevent runaway conversations.

What types of automated tasks can I execute with a Claude CLI agent?▼

The Claude CLI agent supports four automated task types: qa for testing, code-review for examining code, explore for investigation, and implement for building features, each running with deterministic behavior.

Can I run Claude agent tasks in the background and get deterministic results?▼

Yes, the run-agent CLI supports optional background execution for Claude agent tasks while enforcing deterministic behavior through disabled session persistence and prompt-driven task specifications.

How do I prevent runaway conversations when automating Claude agent executions?▼

You prevent runaway conversations by using the run-agent CLI, which enforces strict turn limits, restricts allowed tools, applies hook protection, and disables session persistence for controlled executions.

What is the best way to restrict tools and enforce limits on a Claude agent?▼

The best way to restrict tools and enforce limits is using the run-agent CLI, which applies hook protection, enforces max turns, disables session persistence, and restricts allowed tools for deterministic behavior.

Does the run-agent CLI support session persistence for repeatable agent tasks?▼

No, the run-agent CLI enforces no-session-persistence to ensure repeatable and deterministic results across all automated Claude agent executions, including qa, code-review, explore, and implement tasks.

Related Skills

Explore More Software Engineering Skills→
alexjxalexjx

agent-creator

Create Claude Code subagents as Markdown files with YAML frontmatter.

Community
Basic
hochoa13hochoa13

blackbox

Delegate coding tasks to the Blackbox AI CLI for generation, refactoring, and review.

Community
Advanced
xotongxotong

Hooks Automation

Coordinate Claude Code operations with pre- and post-operation hooks.

Community
Advanced
helix4uhelix4u

llama-cpp

Run LLM inference on CPU, Apple Silicon, and non-NVIDIA GPUs with GGUF models.

Community
Intermediate