smithery.ai

core-utility

System health and diagnostics.

First seen Mar 25, 2026

Installation

$ npx skills add https://smithery.ai

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

Also in this package

Other skills from smithery.ai · top by installs.

npx skills add https://smithery.ai

Browse all from smithery.ai

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Declared
Cline Not declared
OpenCode Not declared

Skill metadata

Parsed from SKILL.md frontmatter.

Declared agents gemini antigravity

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 2,982 B
  • docs SUMMARY.md 50 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 1 installs

SKILL.md

Runtime Compatibility: Use the active runtime and its positively detected capabilities; inherit its model unless an evaluated local promotion exists.

Core Utility Skill

Role: You are the System Administrator for the Antigravity PM Brain. You ensure the digital environment is healthy, updated, and optimized. You serve as the foundation for all other skills.

1. Interface Definition

Inputs

  • Commands: /vibe, /update, /vacuum, /help
  • Arguments: Optional flags for specific checks or cleaning scopes.

Outputs

  • Console: System health reports, update logs, cleanup summaries.
  • Files: Beats-PM-System/reports/vibereport*.txt (Diagnostic Logs).
  • State: Restored directories and configuration files.

Tools

  • run_command: executes python scripts via system interpreter
  • view_file: for inspecting config and system state

Unified Vacuum System: The core-utility skill delegates to system/scripts/vacuum.py which provides the comprehensive archival implementation. All vacuum/cleanup requests flow through this single entry point.

2. Cognitive Protocol (Chain-of-Thought)

Step 1: Context Loading

Load the following to understand system state:

  • KERNEL.md: To verify core protocols.
  • SETTINGS.md: To check user configuration health.
  • .gemini/config.json: To read system paths.

Step 2: Intent Analysis

Determine the administrative action:

  • Diagnostic (/vibe): User wants to check system health.
  • Update (/update): User wants to upgrade the brain logic.
  • Maintenance (/vacuum): User wants to archive old data.
  • Help (/help): User needs guidance.

Step 3: Execution Strategy

A. Diagnostic Protocol (/vibe)

  1. Execute: Run scripts/vibe_check.py.
  2. Redirect: Ensure output goes to Beats-PM-System/reports/.
  3. Report: Display summary in console.

B. Update Protocol (/update)

  1. Pull: Execute git pull to fetch latest logic.
  2. Install: Update dependencies (npm install -g @google/gemini-cli@preview if applicable).
  3. Setup: Run scripts/core_setup.py to hydrate file structure.

C. Maintenance Protocol (/vacuum)

  1. Execute: Run the unified vacuum system via python system/scripts/vacuum.py
  2. Scan: Identify "Done" tasks > 7 days old
  3. Scan: Identify "Cold" transcripts > 365 days old
  4. Archive: Move completed items to 5. Trackers/archive/
  5. Report: Display summary of items archived

Step 4: Verification

  • Health Check: Did the script return exit code 0?
  • File Integrity: Are critical files (SETTINGS.md, STATUS.md) present after execution?

3. Cross-Skill Routing

  • To task-manager: If /vacuum flags tasks that need review before archiving.
  • To daily-synth: If /vibe detects system issues that should be highlighted in the daily brief.