prompt caching

7 articles · 15 co-occurring · 0 contradictions · 0 briefs

Prompt caching is documented as core feature for persistence and efficiency across sessions

Related concepts

tool integration patterns 3 prompt engineering 3 context window management 3 multi agent orchestration 2 memory persistence 2 context window optimization 2 tool workflow patterns 1 tool use architecture 1 token optimization 1 state management 1 skills based tool definition 1 session state management 1 retrieval augmented generation 1 prompt optimization 1 prompt engineering discipline 1

Evidence chain (7 articles, showing 7)

Tool use with Claude - Claude API Docs example_of

Prompt caching is documented as core feature for persistence and efficiency across sessions

Stephen Weber - BLACKBIRD.AI | LinkedIn example_of

Article title 'LLM Prompt Caching: The Hidden Lever' directly addresses context window optimization—a core context engineering pattern.

@sarahwooders: Another reason why skills > MCP: it's much more cache friendly, since adding ... extends

Reveals that prompt caching efficiency is a constraint that should inform architectural decisions about tool integration. Most discussions of prompt caching don't surface this MCP-specific penalty.

@Hesamation: this is a great article if you want to understand Claude Code or Codex and th... example_of

Tweet explicitly names 'prompt cache: stable prefix, only session state changes' as critical component 2. This is direct application of prompt caching to preserve context efficiency.

@Prince_Canuma: If you agent harnesses like Pi by my buddy @badlogicgames, then this mlx-vlm ... extends

Adds persistence layer to prompt caching—moves cached computations to disk rather than losing them between sessions

What is In-Context Learning (ICL)? | IBM example_of

Page lists 'Prompt caching' as a sub-topic under Prompt Optimization, indicating IBM recognizes context window management as important.

@testingcatalog: Anthropic is testing the Insights feature for its Managed Agents on Claude Co... example_of

Summaries and findings are cached so the Insights page loads instantly—cached context is reused across views.

query this concept

$ db.articles("prompt-caching")

$ db.cooccurrence("prompt-caching")

$ db.contradictions("prompt-caching")