Home / Articles / The tokens hook / Versions / v1.1.0
The tokens hook: what changed in v1.1.0
From v1.0.0 (2026-10-15, e4a2670ec) to v1.1.0 (2026-10-18, this release), paragraph by paragraph.
0 paragraphs added, 0 removed, 2 changed in place, 48 unchanged. About 2 words added and 0 removed. Insertions are marked like this, deletions like this; unchanged runs are folded to one line; figures appear as their file names.
26 unchanged paragraphs, under In short {#in-short}, Why tokens need a policy {#why}, Asking the model does not work {#prompt}…
The policy carries the version, the mode, the limits and the prices, with their source. Prices are per model, from Anthropic's pricing page: a five-minute cache write is 1.25 times the base input price, a one-hour write twice, a read a tenth, and "Cache hits and refreshes on Claude Opus 5.5 and Claude Sonnet 5.5 are priced at 0.05x the base input price." The modes are the ones I wantwould like for every policy: observe shows and logs; warn also warns when a limit is crossed; enforce blocks the next prompt once the session's budget is spent.
11 unchanged paragraphs, under The other places to look {#elsewhere}, Other tools {#tools}, A site for one small hook {#site}
I wantwould like this to live at tokens.hook.sgit.ai, as an example of how a mission-critical hook should be run: a small code base, mapped completely, measured, and released like any other software. Hooks are kernel code made the argument; this is the first hook to get the treatment.
11 unchanged paragraphs, under What this does not claim {#limits}, Where this comes from {#sources}