Name: prompt-caching
Rating: 32.800000000000004 (1 reviews)
Author: administrakt0r

prompt-caching

You're a caching specialist who has reduced LLM costs by 90% through strategic caching. You've implemented systems that cache at multiple levels: prompt prefixes, full responses, and semantic similarity matches.

Quality

27%

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

Fix and improve this skill with Tessl

tessl review fix ./plugins/antigravity-bundle-llm-application-developer/skills/prompt-caching/SKILL.md

The canonical home for this skill is prompt-caching in administrakt0r/AI-Agents-Safe-Coding-Skills

Quality

Content

18%Scale 1-5

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

This skill is essentially a skeleton or outline with no actionable content. It lists caching concepts and patterns at a surface level but provides zero executable code, no concrete implementation steps, and no real workflow. The content appears truncated (core principle #2 is cut off) and several sections (anti-patterns) are empty headers, making it unusable as a skill guide.

Suggestions

Add concrete, executable code examples for each pattern (e.g., Anthropic API calls with cache_control blocks, response caching with Redis/dict, CAG implementation with preloaded context).

Define a clear multi-step workflow for implementing prompt caching, including validation steps to verify cache hits are occurring and measuring cost savings.

Fill in the anti-pattern sections with specific examples of what goes wrong and how to avoid it, rather than leaving them as empty headers.

Remove the persona description and 'When to Use' tautology, and replace with actionable decision criteria for choosing between caching strategies.

Dimension	Reasoning	Score
Conciseness	The skill includes some unnecessary persona framing ('You're a caching specialist who has reduced LLM costs by 90%') and explains concepts Claude likely already knows (e.g., 'LLM caching is different from traditional caching'). However, it's not excessively verbose—the tables and lists are reasonably compact. The 'When to Use' section is a meaningless tautology that wastes tokens.	3 / 5
Actionability	The skill provides no concrete code, commands, or executable guidance whatsoever. Every section is a vague label or abstract description—'Use Claude's native prompt caching for repeated prefixes' and 'Cache full LLM responses for identical or similar queries' are descriptions, not instructions. There are no code examples, no API calls, no configuration snippets, and no specific implementation steps.	1 / 5
Workflow Clarity	There is no workflow, no sequenced steps, and no validation checkpoints. The content is a collection of labels and brief descriptions with no process to follow. The core principles list is truncated (item 2 is cut off at 'K'), and the patterns section provides only one-line summaries with no steps.	1 / 5
Progressive Disclosure	The content has some section structure (Patterns, Anti-Patterns, Sharp Edges, Related Skills), but there are no bundle files or references to deeper content. The anti-patterns sections are empty headers with no content. The structure suggests organization but delivers very little substance at any level.	2 / 5
	Total	7 / 20 Passed

Description

36%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.

This description uses first-person voice ('You're a caching specialist who has...') which violates the third-person requirement and reads like a persona prompt rather than a skill description. It lacks a 'Use when...' clause, provides no concrete actions the skill performs, and relies on self-promotional claims ('reduced LLM costs by 90%') rather than describing capabilities and trigger conditions.

Suggestions

Rewrite in third person with concrete actions, e.g., 'Implements multi-level caching strategies for LLM API calls, including prompt prefix caching, response caching, and semantic similarity matching.'

Add an explicit 'Use when...' clause with natural trigger terms, e.g., 'Use when the user asks about reducing LLM/API costs, caching prompts or responses, optimizing token usage, or implementing semantic caching.'

Remove the self-promotional claim ('reduced LLM costs by 90%') and replace with specific capabilities that help Claude distinguish this skill from general optimization skills.

Dimension	Reasoning	Score
Specificity	Names the domain (caching for LLMs) and mentions some concepts like 'prompt prefixes, full responses, semantic similarity matches,' but these read more like buzzwords than concrete actions the skill performs. No clear verbs describing what the skill actually does beyond vague 'implemented systems.'	2 / 5
Completeness	There is a vague 'what' (caching at multiple levels) but no 'when' clause at all. There is no explicit guidance on when Claude should select this skill. The description reads as a resume bullet point rather than a skill selector.	2 / 5
Trigger Term Quality	Includes some relevant keywords like 'caching,' 'LLM costs,' 'prompt prefixes,' and 'semantic similarity,' but misses natural user phrases like 'reduce API costs,' 'cache responses,' 'token savings,' or 'API optimization.' The terms are more self-promotional than search-friendly.	3 / 5
Distinctiveness Conflict Risk	The focus on LLM caching specifically is somewhat distinctive, but the broad mention of 'reducing costs' and 'caching' could overlap with general optimization or cost-reduction skills. The lack of concrete boundaries increases conflict risk.	3 / 5
	Total	10 / 20 Passed

Validation

90%

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 — 10 / 11 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	10 / 11 Passed

Repository: administrakt0r/AI-Agents-Safe-Coding-Skills
Path: plugins/antigravity-bundle-llm-application-developer/skills/prompt-caching/SKILL.md
Commit: 57c135b

Reviewed: 3 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.