Does it summarize the output of the command?

No, unlike standard execution, this skill is specifically instructed to return the full, raw output from the terminal without any summarization.

Why use this instead of standard Bash execution?

It provides a 'strict mode' for terminal interactions, preventing the AI from being 'helpful' by adding flags like --stat or launching subagents during execution.

Does this skill modify my terminal commands?

No, the primary purpose of the runexact skill is to execute the command exactly as provided without any added flags, interpretation, or syntax changes.

Can I use this for complex Bash scripts?

Yes, it is ideal for complex strings, pipes, and multi-part commands that an AI might otherwise try to simplify or rewrite incorrectly.

Which tool does this skill utilize?

It utilizes the built-in Bash tool but applies a restrictive policy to ensure the command is passed through without any modification.

Exact Command Runner

Name: Exact Command Runner
Author: AI-Builder-Team

byAI-Builder-Team

•

开发者工具

Executes terminal commands verbatim without modification or output summarization to ensure precise control over shell interactions.

The Exact Command Runner skill ensures that terminal commands are executed precisely as written, bypassing Claude's tendency to add flags, interpret intent, or truncate long outputs. This skill is indispensable when specific syntax is required for legacy systems, sensitive script execution, or when raw, unedited output is necessary for debugging or downstream processing. By enforcing a strict verbatim execution policy, it provides developers with a deterministic way to interact with the shell without AI-driven alterations like the automatic addition of summary flags or subagent interference.

主要功能

01Prevention of automatic flag injection (like --stat)

02Deterministic single-call execution

03Verbatim command execution via Bash tool

042 GitHub stars

05Zero-modification policy for input strings

06Full, unsummarized output retrieval

使用场景

01Running scripts where specific flags are sensitive and must not be altered by the model

02Executing complex piping or redirection commands that AI might otherwise simplify

03Capturing entire logs or process outputs for verification without AI-driven truncation

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ai-builder-team/ai-builder-plugin-marketplace runexact

For use in Claude.ai and ChatGPT

Download Skill