{"version":"2.1.216","anchor":"context-over-limit-warning-improved","canonical_anchor":"context-over-limit-warning-improved","heading":"Context Over-Limit Warning Improved","tier":null,"area":null,"scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.216\/e\/context-over-limit-warning-improved","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.216","markdown":"### Context Over-Limit Warning Improved\n\n**What**\n\nThe message shown when the context window is exhausted now distinguishes between the auto-compaction window and the raw model limit, and always names the specific action to take.\n\n**Details**\n\n- Auto mode: `Context exceeds the N-token limit by M tokens \u2014 run \/compact or \/clear to continue.`\n- Non-auto mode: `Context is M tokens past the N-token compaction window \u2014 run \/compact to reduce usage.`\n- If `DISABLE_COMPACT` is set: suggests only `\/clear`\n\n**Evidence**\n\nContext limit message (search for `\"Context exceeds the\"` or `\"Context is\"` combined with `\"token compaction window\"`)."}