mixed-precision-trainer
Mixed Precision Trainer - Auto-activating skill for ML Training. Triggers on: mixed precision trainer, mixed precision trainer Part of the ML Training skill category.
36
Quality
3%
Does it follow best practices?
Impact
96%
1.07xAverage score across 3 eval scenarios
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./planned-skills/generated/07-ml-training/mixed-precision-trainer/SKILL.mdQuality
Discovery
7%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This description is severely underdeveloped, essentially serving as a placeholder rather than a functional skill description. It lacks any concrete capabilities, has redundant and narrow trigger terms, and provides no guidance on when Claude should select this skill. The only redeeming quality is that 'mixed precision' is a somewhat distinctive technical term.
Suggestions
Add specific capabilities: 'Configures automatic mixed precision (AMP) training, manages gradient scaling, optimizes GPU memory usage for FP16/BF16 training.'
Expand trigger terms to include natural variations: 'FP16', 'half precision', 'AMP', 'automatic mixed precision', 'gradient scaling', 'memory optimization', 'CUDA memory'.
Add explicit 'Use when...' clause: 'Use when the user asks about reducing GPU memory usage during training, enabling half-precision training, or configuring AMP in PyTorch/TensorFlow.'
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description only names the skill ('Mixed Precision Trainer') and its category ('ML Training') without describing any concrete actions or capabilities. No specific operations like 'configure FP16 training', 'manage gradient scaling', or 'optimize memory usage' are mentioned. | 1 / 3 |
Completeness | The description fails to answer 'what does this do' beyond the name, and the 'when' clause is just a repetition of the skill name rather than meaningful trigger guidance. No explicit 'Use when...' clause with actionable triggers. | 1 / 3 |
Trigger Term Quality | The trigger terms are redundant ('mixed precision trainer' listed twice) and overly narrow. Missing natural variations users might say like 'FP16', 'half precision', 'AMP', 'automatic mixed precision', 'gradient scaling', or 'memory optimization'. | 1 / 3 |
Distinctiveness Conflict Risk | The term 'mixed precision' is fairly specific to a particular ML technique, which provides some distinctiveness. However, the generic 'ML Training' category and lack of specific capabilities could cause overlap with other training-related skills. | 2 / 3 |
Total | 5 / 12 Passed |
Implementation
0%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill is an empty template with no actual content about mixed precision training. It contains only meta-descriptions of what the skill should do without any concrete implementation details, code examples, or actionable guidance. The content provides zero value as it teaches Claude nothing about mixed precision training techniques, PyTorch/TensorFlow APIs, or best practices.
Suggestions
Add executable code examples showing mixed precision training setup (e.g., PyTorch's torch.cuda.amp.autocast() and GradScaler, or TensorFlow's mixed_precision.set_global_policy())
Include specific workflow steps: 1) Check GPU compatibility, 2) Configure precision policy, 3) Wrap forward pass, 4) Scale gradients, 5) Validate training stability
Provide concrete guidance on common pitfalls like loss scaling, gradient overflow detection, and which layers should remain in full precision
Add a quick-start code snippet that demonstrates a minimal working mixed precision training loop
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is entirely filler with no actual technical information. It explains what the skill does in abstract terms without providing any concrete guidance, wasting tokens on meta-descriptions Claude doesn't need. | 1 / 3 |
Actionability | No executable code, commands, or specific instructions are provided. The content only describes capabilities in vague terms like 'provides step-by-step guidance' without actually providing any guidance. | 1 / 3 |
Workflow Clarity | No workflow, steps, or process is defined. The skill claims to provide 'step-by-step guidance' but contains zero actual steps for implementing mixed precision training. | 1 / 3 |
Progressive Disclosure | No references to detailed documentation, no code examples to link to, and no structured content organization. The entire file is a placeholder with no substantive content to organize. | 1 / 3 |
Total | 4 / 12 Passed |
Validation
81%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 9 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
allowed_tools_field | 'allowed-tools' contains unusual tool name(s) | Warning |
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 9 / 11 Passed | |
- Commit
f17dd51
- Reviewed
Table of Contents
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.