mirror of
https://github.com/discourse/discourse.git
synced 2026-08-06 13:08:40 +08:00
f86bdd738aApply token-budgeted agent execution and context compaction to all AI agent runs instead of gating it behind execution_mode. Remove the legacy fixed-limit agent settings, keep compression_threshold defaulted to 80, and trim prompt history from the latest compression checkpoint. Why: Preserving a stable compressed context across turns keeps important conversation state available while allowing newer messages to append to a consistent prefix. That improves cache reuse and avoids repeatedly throwing away useful context just to stay under model limits. Compaction shape: before: turn 1 [full -> compact] | turn 2 [full -> compact] after: [compact checkpoint] -> turn 1 -> turn 2 -> ... --------- Co-authored-by: Rafael Silva <xfalcox@gmail.com> |
||
|---|---|---|
| .. | ||
| admin/assets/javascripts/admin | ||
| app | ||
| assets | ||
| config | ||
| db | ||
| lib | ||
| spec | ||
| test/javascripts | ||
| package.json | ||
| plugin.rb | ||
| README.md | ||
| tsconfig.json | ||
Data Explorer Plugin
This plugin allows admins to run SQL queries against the live Discourse database, including parameterized queries and formatting for several common column types.
For more information, please see: https://meta.discourse.org/t/data-explorer-plugin/32566
