🎯

prompt-caching

🎯Skill

from hainamchung/agent-assistant

What it does

Reduces LLM costs by strategically caching prompts, responses, and semantic matches across multiple caching levels.

📦

Part of

hainamchung/agent-assistant(227 items)

prompt-caching

Installation

npm installInstall npm package

npm install -g @namch/agent-assistant

git cloneClone repository

git clone https://github.com/hainamchung/agent-assistant.git

Node.jsRun Node.js server

node cli/install.js install cursor # Cursor

Node.jsRun Node.js server

node cli/install.js install claude # Claude Code

Node.jsRun Node.js server

node cli/install.js install copilot # GitHub Copilot

+ 7 more commands

📖 Extracted from docs: hainamchung/agent-assistant

Need more details? View full documentation on GitHub →

1Installs

AddedFeb 4, 2026

View on GitHub Back to Skills

Skill Details

SKILL.md

"Caching strategies for LLM prompts including Anthropic prompt caching, response caching, and CAG (Cache Augmented Generation) Use when: prompt caching, cache prompt, response cache, cag, cache augmented."

Overview

# 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.

You understand that LLM caching is different from traditional caching—prompts have

prefixes that can be cached, responses vary with temperature, and semantic similarity

often matters more than exact match.

Your core principles:

Cache at the right level—prefix, response, or both
K

Capabilities

prompt-cache
response-cache
kv-cache
cag-patterns
cache-invalidation

Patterns

Anthropic Prompt Caching

Use Claude's native prompt caching for repeated prefixes

Response Caching

Cache full LLM responses for identical or similar queries

Cache Augmented Generation (CAG)

Pre-cache documents in prompt instead of RAG retrieval

Anti-Patterns

❌ Caching with High Temperature

❌ No Cache Invalidation

❌ Caching Everything

⚠️ Sharp Edges

| Issue | Severity | Solution |

|-------|----------|----------|

| Cache miss causes latency spike with additional overhead | high | // Optimize for cache misses, not just hits |

| Cached responses become incorrect over time | high | // Implement proper cache invalidation |

| Prompt caching doesn't work due to prefix changes | medium | // Structure prompts for optimal caching |

Related Skills

Works well with: context-window-management, rag-implementation, conversation-memory

More from this repository10

Skill

Develops high-performance C++ applications with modern C++20/23 features, template metaprogramming, and zero-overhead systems design.

🎯

senior-architect🎯Skill

Designs scalable software architectures using modern tech stacks, generating architecture diagrams, analyzing dependencies, and providing system design recommendations.

🎯

senior-frontend🎯Skill

Generates, analyzes, and scaffolds modern frontend projects using ReactJS, NextJS, TypeScript, and Tailwind CSS with automated best practices.

🎯

spec-miner🎯Skill

Extracts and documents specifications from legacy or undocumented codebases by systematically analyzing code structure, data flows, and system behaviors.

🎯

docs-seeker🎯Skill

Searches and retrieves technical documentation by executing intelligent scripts across library sources, GitHub repos, and context7.com with automated query detection.

🎯

writing-plans🎯Skill

Generates comprehensive, step-by-step implementation plans for software features with precise file paths, test-driven development approach, and clear task granularity.

🎯

file path traversal testing🎯Skill

Tests and identifies potential file path traversal vulnerabilities in code by analyzing file path handling and input validation mechanisms.

🎯

nodejs-best-practices🎯Skill

Guides developers in making strategic Node.js architecture and framework decisions by providing context-aware selection principles and modern runtime considerations.

🎯

red-team-tactics🎯Skill

Simulates adversarial attack techniques across MITRE ATT&CK framework phases, mapping network vulnerabilities and demonstrating systematic compromise strategies.