veomni-uv-update

Use this skill when updating dependencies managed by uv: bumping a package version, upgrading the uv tool itself, updating torch/CUDA stack, switching transformers version, or regenerating the lockfile. Trigger: 'update dependency', 'bump version', 'upgrade uv', 'update torch', 'update lockfile', 'uv sync fails'.

Quality

98%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

96%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

An exceptionally actionable, concise operational guide with clear sequenced workflows, validation checkpoints, and a pitfalls checklist for fragile batch dependency updates. Its only weakness is progressive disclosure: nearly all detail is inlined in one long file with a single external pointer, so it could benefit from splitting the torch/attention-kernel breakdown into a reference file.

Suggestions

Move the detailed Scenario 3 attention-kernel taxonomy (prebuilt wheels / PyPI releases / source-built git pins) into a references/ file (e.g. references/torch-wheels.md) and summarize the three groups inline, reducing SKILL.md length and improving progressive_disclosure.

Add a short 'Verification' checklist at the end of each destructive scenario (lockfile diff clean, tests green, Dockerfile grep shows no drift) to make the existing validation steps uniform and easy to scan.

Confirm the referenced '.agents/knowledge/uv.md' exists in the repo and consider linking it as a bundle reference so the dependency-architecture pointer is verifiable rather than implicit.

Dimension	Reasoning	Score
Conciseness	Dense and project-specific throughout — standalone Dockerfiles, direct wheel URLs, three hardware extras, override markers — with no padding or explanation of concepts Claude already knows; every line carries non-obvious VeOmni-specific knowledge, matching the level-5 'every token earns its place' anchor.	5 / 5
Actionability	Provides copy-paste-ready commands — 'grep -rn "astral-sh/uv" docker/', 'uv lock', 'uv sync --extra gpu --dev', 'pytest tests/' — plus exact pyproject paths and upstream URLs, fully executable across the common cases, matching the level-5 anchor.	5 / 5
Workflow Clarity	Each scenario is a numbered sequence with explicit validation checkpoints ('git diff uv.lock — should only show version changes', 'pytest tests/') and a feedback loop ('If uv lock fails... check conflicts, override markers, wheel availability'); the Common Pitfalls section acts as a checklist. Validation is present for the batch/destructive operations, so the level-3 cap does not apply, and it reaches the level-5 anchor.	5 / 5
Progressive Disclosure	Well-organized into 'Before You Start' overview plus Scenarios 1–5 and Common Pitfalls, with one clearly-signaled one-level reference ('Read .agents/knowledge/uv.md for the full dependency architecture'). It falls short of level-5 because the detailed content is not split across bundle files — everything is inlined in a single ~170-line file rather than offloaded to references, and no references/scripts/assets bundle exists to verify.	4 / 5
	Total	19 / 20 Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, well-scoped description: it states exactly what the skill does, when to invoke it, and supplies natural trigger phrases with a failure-mode hook. Specific, complete, and clearly distinct from sibling skills.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'bumping a package version, upgrading the uv tool itself, updating torch/CUDA stack, switching transformers version, or regenerating the lockfile' — giving comprehensive coverage of the skill's scope, matching the level-5 anchor rather than the level-4 'minor gaps' anchor.	5 / 5
Completeness	Explicitly answers both 'what' (updating uv-managed dependencies with enumerated actions) and 'when' (an explicit 'Use this skill when' clause plus concrete trigger phrases), matching the level-5 anchor; not level-4 because the 'when' is fully explicit with named triggers.	5 / 5
Trigger Term Quality	Provides a Trigger list of natural phrases users would say — 'update dependency', 'bump version', 'upgrade uv', 'update torch', 'update lockfile', 'uv sync fails' — covering synonyms and a failure-mode trigger; comprehensive enough for the level-5 anchor.	5 / 5
Distinctiveness Conflict Risk	The 'dependencies managed by uv' niche plus torch/CUDA/transformers/lockfile specifics carve a clear, distinct trigger surface with minimal overlap risk against generic dependency or build skills, matching the level-5 anchor.	5 / 5
	Total	20 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
referenced_paths_exist	Referenced path issues: 1 missing, 1 deeper-than-1-level	Warning

	Total	15 / 16 Passed

Repository: ByteDance-Seed/VeOmni
Path: .agents/skills/veomni-uv-update/SKILL.md
Commit: b8a3edc

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.