lev-os/agents/ubs

Claude Code and agent skill by lev-os/agents

Trust tier 2
  • community
  • ext:[{"type":"hook","description":"Pre/post lifecycle hooks"}]
LLM review pending
Source check pending

Checking that the SKILL.md file is still in the repository.

[WHAT] Static-analysis quality-gate wrapper around the external `ubs` CLI (Ultimate Bug Scanner). [HOW] Runs `ubs` on staged, diffed, or explicitly changed files with structured output, then uses the findings as a pre-commit or pre-review bug gate. [WHEN] Use before commit, during code review, when the user asks for bug scanning, or when a change needs a fast multi-language sanity pass. [WHY] Catches AI-agent bug patterns early without making UBS itself the project workflow router. Triggers: "ubs", "ultimate bug scanner", "bug scan", "static analysis", "quality gate", "scan for bugs", "pre-commit scan"

Install

npx vskill@latest install lev-os/agents/ubs

Semi-extensibleMentions customization, but needs skill-specific knowledge. Read the extensible skills standard

SKILL.md

UBS

Use ubs as a scoped gate. Keep the scope tight unless the user explicitly wants a full baseline scan.

Prerequisites

bash
command -v ubs

If missing, stop and tell the user:

bash
brew install dicklesworthstone/tap/ubs

Golden Rules

  1. Prefer --staged, --diff, or explicit files over ubs ..
  2. Use --format=json when you need machine-readable output.
  3. Re-run after fixes to confirm the exit code is clean.
  4. Treat UBS as a tool boundary; do not import its repo-level hooks as implicit Lev policy.

Fast Path

bash
ubs --staged --format=json --ci

If you only changed a few files:

bash
ubs path/to/file.ts path/to/other.py --format=json --ci

Core Commands

Narrow scans
bash
ubs --staged --format=json --ci
ubs --diff --format=json --ci
ubs src/file.ts src/other.py --format=json --ci
Full or filtered scans
bash
ubs . --format=json --ci
ubs --only=js,python . --format=json --ci
ubs --category=resource-lifecycle . --format=json --ci
Diagnostics
bash
ubs doctor
ubs doctor --fix

Recommended Workflow

Before commit
bash
ubs --staged --fail-on-warning --format=json --ci
During review
  1. Scan only the touched files.
  2. Inspect critical findings first.
  3. Fix root causes, not symptoms.
  4. Re-run the same narrow scan.

Output Handling

Focus on:

  • summary
  • findings[].file
  • findings[].line
  • findings[].severity
  • findings[].category
  • findings[].message

If the user wants GitHub or CI artifacts:

bash
ubs . --format=sarif --ci
ubs . --report-json .ubs/latest.json --html-report .ubs/latest.html --ci

When Not To Use UBS

Do not default to UBS when:

  • the user wants architectural review rather than static findings
  • there are no changed files and no request for a baseline scan
  • another domain-specific checker is clearly more appropriate

Anti-Patterns

  • Running ubs . on every small edit
  • Ignoring critical findings without checking the code
  • Treating UBS findings as proof without context
  • Installing repo hooks or settings from the upstream repo without explicit user approval

Minimal Report Pattern

markdown
## UBS Findings
-Scope scanned:
-Critical:
-Warnings:
-Follow-up fixes:

Keep the summary in the active Lev artifact or review comment, not in detached scratch notes.

Security

  • Security scanPassed, 100 of 100Pattern checks on the skill's files.
  • Model reviewNot runA model review is scheduled for the next rescan.

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

Repository stars
23
Repository forks
2
7-day trend score
21

Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.

Extension points

  • hookPre/post lifecycle hooks

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/lev-os/agents/ubs/badge)](https://verified-skill.com/skills/lev-os/agents/ubs)

Something wrong with this skill or its page? Tell us.