Restrict file edits and block destructive commands during live debugging.
npx skills add https://github.com/bhianerd/aimemory --skill guard-bhianerd
Please help me install this Agent Skill. Skill: guard Source: https://github.com/bhianerd/aimemory/tree/main/.claude/skills/gstack/guard Command: npx skills add https://github.com/bhianerd/aimemory --skill guard-bhianerd
Prevents accidental destructive actions and risky edits by combining destructive-command warnings with a directory-scoped edit boundary, which reduces mistakes during production work and live debugging.
Ask the user to run /guard, then provide the directory path you want to allow edits within when prompted.
High-intent search queries and answers about installing and using this skill.
Preventing accidental destructive commands during production debugging requires a combined guard mode that flags high-risk operations like rm -rf or DROP TABLE before execution. It restricts edits to a specific directory and blocks unsafe command execution.
A directory-scoped edit boundary is a freeze mechanism that blocks writing or editing files outside a user-selected absolute directory path. It protects sensitive repos by limiting accidental out-of-scope modifications during live troubleshooting workflows.
To set up a freeze boundary for live troubleshooting, invoke the guard mode and provide the absolute directory path you want to allow edits within when prompted. The system then enforces directory-scoped restrictions using a trailing slash state file to confine file writes.
Yes, you can restrict file edits to a specific directory when modifying live production code by enabling a freeze boundary. This blocks any writing or editing actions outside your resolved absolute directory path to limit potential damage.
You should use a combined guard mode for live debugging workflows when working in prod-like environments or sensitive repos and need maximum safety. It is ideal for situations requiring both destructive command warnings and directory-scoped edit restrictions simultaneously.
shogo-labsBlock file edits outside a user-specified directory during a session.
Walrus-ComputingWarn before destructive shell commands and restrict file edits to an approved directory.
NaKMiersBlock unsafe operations with destructive command warnings and directory-based edit boundaries.
huichenWarn on destructive commands and restrict edits to a specified directory.
