xonecas.io
64875
Mc

mini-coder / commit 3601cef

fix: reduce model sync TTL from 7 days to 24 hours

New models added to Zen (e.g. gpt-5.4-mini, gpt-5.4-nano) were not
appearing until the 7-day cache expired. A 24-hour TTL catches new
models much sooner.

xonecas2026-03-19 3601cef2a884c84606e85b9d64e36adac92dca3b parent d602753

2 files changed

TODO.md+1−3
@@ -1,8 +1,6 @@
11 # TODO
22
3-- Status bar drops the cwd path when the line gets long (observed with long OAuth model names on second turn). May be a terminal width truncation issue.
4-- Model DB sync doesn't include newer Zen models. `gpt-5.4-nano`, `gpt-5.4-mini`, `gpt-5.4` are available on Zen but absent from `provider_models`. The sync may be running against a stale endpoint or filtering too aggressively.
5-- Need to review our parity with opencode and claude code configs (skills, agents, and commands). Identify gaps and missing functionality.
3+- Need to review our parity with opencode and claude code configs (skills, agents, and commands). Identify gaps and missing functionality. use online docs for both to assert functionality
64 - CTRL+d hangs sometimes and doesn't exit.
75 - Codebase is hurting from fast iteration and adding/removing features as well as core fundamental changes.
86
src/llm-api/model-info.ts+1−1
@@ -34,7 +34,7 @@ const MODELS_DEV_SYNC_KEY = "last_models_dev_sync_at";
3434 const PROVIDER_SYNC_KEY_PREFIX = "last_provider_sync_at:";
3535 const CACHE_VERSION_KEY = "model_info_cache_version";
3636 const CACHE_VERSION = 3;
37-export const MODEL_INFO_TTL_MS = 7 * 24 * 60 * 60 * 1000;
37+export const MODEL_INFO_TTL_MS = 24 * 60 * 60 * 1000;
3838
3939 export {
4040 buildModelMatchIndex,