prompt-repetition
A prompt repetition technique for improving LLM accuracy. Achieves significant performance gains in 67% (47/70) of 70 benchmarks. Automatically applied on lightweight models (haiku, flash, mini).
65
47%
Does it follow best practices?
Impact
97%
1.56xAverage score across 3 eval scenarios
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./.agent-skills/prompt-repetition/SKILL.mdQuality
Discovery
17%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
The description provides some technical specificity about the technique and its effectiveness but fails to include natural trigger terms users would say and lacks explicit guidance on when Claude should select this skill. The automatic application mention is helpful but doesn't address manual invocation scenarios.
Suggestions
Add a 'Use when...' clause specifying trigger scenarios, e.g., 'Use when the user requests improved accuracy, more reliable responses, or when working with smaller/faster models'
Include natural language trigger terms users would actually say, such as 'better answers', 'more accurate', 'improve quality', 'reliable responses'
Clarify what concrete actions the skill performs beyond automatic application, e.g., 'Repeats key instructions in prompts to reduce errors and improve response quality'
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Names the domain ('prompt repetition technique for improving LLM accuracy') and mentions specific outcomes ('67% of 70 benchmarks', 'lightweight models'), but doesn't list concrete actions the skill performs beyond 'automatically applied'. | 2 / 3 |
Completeness | Partially explains what it does (prompt repetition for accuracy) but has no explicit 'Use when...' clause. The 'when' is only implied through 'automatically applied on lightweight models' which describes automatic behavior, not user-triggered scenarios. | 1 / 3 |
Trigger Term Quality | Uses technical jargon ('prompt repetition technique', 'LLM accuracy', 'haiku, flash, mini') that users wouldn't naturally say. Missing natural trigger terms like 'improve responses', 'better answers', or 'accuracy'. | 1 / 3 |
Distinctiveness Conflict Risk | The focus on 'prompt repetition' and specific model types (haiku, flash, mini) provides some distinctiveness, but 'improving LLM accuracy' is broad enough to potentially overlap with other optimization or enhancement skills. | 2 / 3 |
Total | 6 / 12 Passed |
Implementation
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a comprehensive, highly actionable skill with excellent executable code and clear workflows. However, it suffers from verbosity—explaining causal attention mechanisms and including extensive research details that could be referenced externally. The content would benefit from splitting into a concise SKILL.md overview with detailed implementation and research in separate files.
Suggestions
Move the 'How It Works' causal attention explanation to a separate THEORY.md file—Claude understands transformer architecture
Extract the production-ready implementation code to a separate IMPLEMENTATION.md or actual Python file, keeping only a brief usage example in SKILL.md
Condense the research results section to key findings (67% improvement rate, 0 degradation) and link to a RESEARCH.md for full benchmark details
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill contains valuable information but is verbose in places, explaining concepts like causal attention that Claude likely understands. The 500+ line document could be condensed significantly while preserving actionable content. | 2 / 3 |
Actionability | Provides fully executable Python code, concrete examples with before/after prompts, specific model lists, and copy-paste ready implementations. The transformer class and integration examples are production-ready. | 3 / 3 |
Workflow Clarity | Clear 4-step application procedure with explicit decision points, token limit checks, and validation. The A/B testing method provides verification steps, and the multi-agent integration section shows clear sequencing with duplicate prevention. | 3 / 3 |
Progressive Disclosure | Content is well-organized with clear sections and a quick reference at the end, but the document is monolithic at 500+ lines. The production implementation, examples, and research details could be split into separate reference files. | 2 / 3 |
Total | 10 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
skill_md_line_count | SKILL.md is long (544 lines); consider splitting into references/ and linking | Warning |
Total | 10 / 11 Passed | |
- Repository
- supercent-io/skills-template
- Commit
c033769
- 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.