rag-forge cache

Semantic cache management

Synopsis

rag-forge cache <subcommand>

Description

cache provides visibility into the semantic query cache. The semantic cache intercepts incoming queries and returns stored answers for semantically similar previous questions, avoiding redundant LLM and embedding calls and directly reducing the costs reported by rag-forge cost.

Subcommands

cache stats

Print hit/miss statistics for the semantic cache.

rag-forge cache stats

The output shows the hit rate as a percentage, the raw hit and miss counts, the total number of queries recorded, and the cache backend in use. No options are available — the command reads configuration from the project’s rag-forge.config.ts.

Example

rag-forge cache stats

Related commands

• rag-forge query — run a query (cache hits are recorded here)
• rag-forge cost — estimate pipeline costs (cache hits reduce LLM spend)