Claude's own replies are scanned for text imitating internal control markup, and matches counted only.
What's wrong with this entry?
Main-session assistant messages are now checked for text that looks like Claude Code's own internal control markup, including system reminders, harness envelopes and token-count, session-context, wake and role-turn shaped lines. Nothing is rewritten; the matches are only counted and reported.
- Runs for main-session, non-error assistant messages only, with no flag involved.
tengu_assistant_output_harness_content
Strings lifted out of the shipped bundle, so the claim above can be checked against them.