Skip to content
NOTES
MenuClose
NOTES // rule

Cache-stable prefix

LAST_MODIFIED
2026.07.20
CATEGORY
rule

A prompt cache is a prefix match on exact bytes. Anything that changes above the marker - timestamps, request IDs, reordered tools, dynamic tool results - invalidates every cached token after it. You still get an answer; you pay full price plus the write premium.

Rule

  1. Freeze system prompt + tool schemas before the cache marker.
  2. Put dynamic fields after it.
  3. Never rewrite text that already sits inside a cached block mid-session.
  4. Verify with usage.cache_read_input_tokens on call 2+ of an identical prefix - zero means the prefix is changing or shorter than the model minimum.

Playbook with copy-paste and the measured failure modes: Don’t break the prompt cache. Pair with Trim output, not the cache when the temptation is to compress history in place.

cachingcostprompt-cachingtokens