cctally 1.22.0 → 1.22.1

package/CHANGELOG.md

@@ -5,6 +5,12 @@ based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 
 ## [Unreleased]
 
+## [1.22.1] - 2026-05-30
+
+### Changed
+- **Internal refactor (no user-facing change): extracted the CLI argument-parser construction and the shareable-output construction out of the `bin/cctally` single-file program into two new stdlib-only sibling modules, `bin/_cctally_parser.py` and `bin/_cctally_share.py` (batch 2 of the ongoing `bin/cctally` split).** This trims the main script by ~4,130 lines (16,246 → 12,112) with no change to any command's behavior, flags, output, or exit codes — every one of the 1,329 harness + pytest checks passes unchanged — and both modules ship in the npm/brew/public packages (promoted to the mirror allowlist). Purely a maintainability / code-organization change; nothing to do on upgrade.
+- **Exact embedded pricing for 28 more OpenAI/Codex `gpt-5.x` models, so their costs are computed precisely instead of approximated.** The Codex reports (`codex-daily`/`monthly`/`weekly`/`session`) price every session locally from the embedded `CODEX_MODEL_PRICING` table; models absent from it previously fell back to `gpt-5` pricing (flagged `isFallback`, an approximation). This sync adds the full set of openai-provider `gpt-5*` variants the LiteLLM snapshot lists but the curated table didn't carry — the bare/dated `gpt-5`/`gpt-5.1`/`gpt-5.2` rows plus the `-chat`, `-mini`, `-nano`, `-search-api`, and `-pro` variants and the `gpt-5.4`/`gpt-5.5` families — at their exact LiteLLM rates (the four `*-pro` models, which omit a cached-input rate upstream, use the documented input/4 cache fallback). It also backfills the now-published above-272k tier for `gpt-5.5` and bumps `PRICING_SNAPSHOT_DATE` to 2026-05-30, so `cctally pricing-check` now reports nothing missing. No Claude pricing changed (it was already complete and matched the snapshot exactly).
+
 ## [1.22.0] - 2026-05-29
 
 ### Added