dangerous-manifest_skill

Overview

This skill exposes archived versions of skills from a centralized archive and requires elevated permissions and external binaries to operate. It is intended for controlled backup and forensic workflows where broad filesystem access and native helper tools are acceptable. Use only in trusted environments with strict isolation and monitoring.

How this skill works

The skill locates and reads archived skill data, indexes metadata, and can invoke native helper binaries to unpack or transform archived files. It may request wide filesystem access to collect, validate, and move archived artifacts for backup or analysis. Permission prompts and environment validation are recommended before enabling the skill.

When to use it

• You need to extract or restore archived skill versions from a centralized archive for recovery.
• Performing forensic analysis on archived skill packages in a secure lab environment.
• Automating backups that require native tools to decompress or verify large archived files.
• Migrating archived skill content between isolated systems under administrator control.

Best practices

• Run the skill only on isolated hosts or within short-lived, monitored containers.
• Grant the minimum necessary filesystem and binary access; prefer scoped directories.
• Use signed or checksummed archives and validate binaries before execution.
• Enable audit logging and real-time monitoring of the skill’s actions.
• Review and approve any external binary downloads or environment changes manually.

Example use cases

• Restore a specific historical version of a skill to reproduce an old behavior for debugging.
• Bulk export archived skill metadata and packages to a secure cold-storage backup.
• Analyze archived packages for regressions, deprecated dependencies, or security issues.
• Unpack and repackage archived skills using native decompression and verification tools.