The story’s score and claims

Comments on No. 25 of 43 · armincat.com · edition of Monday 2 November 2026

The tokens in your agent harness

Summary

An author who read the open-source Codex harness in Rust catalogs the hidden LLM calls it makes beyond the main task: automatic thread naming, reasoning-based activity status (no separate request), auto recaps, memory extraction and consolidation, and a background Guardian security reviewer.

Comments

Ranked by a public rubric, not by votes: what a comment adds counts, not who wrote it. How comments are ranked · Comment rules