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'.
90%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Low
Low-risk findings worth noting
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
The required Scenario 3 workflow directs the agent to consult the third-party Luosuu/flash-attention3-wheels GitHub releases page to find matching wheels, so it actively fetches outsider-authored release content without the outsider being able to force delivery into the workflow.
The skill fetches instructions or code from an external URL at runtime, and the fetched content directly controls the agent’s prompts or executes code. This dynamic dependency allows the external source to modify the agent’s behavior without any changes to the skill itself.
The skill is documentation guiding a developer to pick wheel versions from a personal, individual-owned GitHub release page (Luosuu/flash-attention3-wheels) whose matching wheels are then pinned in pyproject.toml and installed by uv, making it a common/reputable unofficial inbound dependency (medium), while the download.pytorch.org reference is official and does not raise the tier.
ByteDance-Seed/VeOmni
35e3bb7
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.