You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: CONFIGURATION.md
+2Lines changed: 2 additions & 0 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -114,6 +114,7 @@ Higher-tier models with longer cache windows benefit from a longer TTL. Setting
114
114
|`cache_ttl`|`string` or `object`|`"5m"`| Time after a response before applying pending ops. String or per-model map. |
115
115
|`protected_tags`|`number` (1–100) |`20`| Last N active tags immune from immediate dropping. |
116
116
|`nudge_interval_tokens`|`number`|`10000`| Minimum token growth between rolling nudges. |
117
+
|`toast_duration_ms`|`number` (1000–60000) |`5000`| TUI toast lifetime for Magic Context notifications in milliseconds. Increase this if toasts disappear too quickly. |
117
118
|`execute_threshold_percentage`|`number` (20–80) or `object`|`65`| Context usage that forces queued ops to execute. Capped at 80% max for cache safety. Supports per-model map. |
118
119
|`execute_threshold_tokens`|`object` (per-model map) | — |**Optional absolute-tokens variant of `execute_threshold_percentage`.** Per-model map (e.g. `{ "default": 150000, "github-copilot/gpt-5.2-codex": 40000 }`). When set for a model, overrides the percentage-based threshold for that model. Clamped to `80% × context_limit` with a warn log. Requires a resolvable context limit — falls through to percentage if unavailable. See below. |
119
120
|`auto_drop_tool_age`|`number`|`100`| Auto-drop tool outputs older than N tags during execution. |
@@ -634,6 +635,7 @@ Tier boundaries are hardcoded to keep behavior predictable and prevent cache-bus
0 commit comments