getcompanion-ai/co-mono
1
0
Fork 0
mirror of https://github.com/getcompanion-ai/co-mono.git synced 2026-04-15 19:05:11 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Clarify messagesToSummarize starts after last compaction

Browse source
This commit is contained in:
Mario Zechner 2025-12-24 12:44:05 +01:00
parent ee0befdfe1
commit 97bbd7a642
1 changed files with 2 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

4
packages/coding-agent/docs/hooks.md
View file

@ -194,8 +194,8 @@ The `before_compact` event lets you implement custom compaction strategies. Unde
|-------|-------------|
| `entries` | All session entries (header, messages, model changes, previous compactions). Use this for custom schemes that need full session history. |
| `cutPoint` | Where default compaction would cut. `firstKeptEntryIndex` is the entry index where kept messages start. `isSplitTurn` indicates if cutting mid-turn. |
| `messagesToSummarize` | Messages that will be summarized and discarded (before cut point). |
| `messagesToKeep` | Messages that will be kept verbatim after the summary (after cut point). |
| `messagesToSummarize` | Messages that will be summarized and discarded (from after last compaction to cut point). |
| `messagesToKeep` | Messages that will be kept verbatim after the summary (from cut point to end). |
| `tokensBefore` | Current context token count (why compaction triggered). |
| `model` | Model to use for summarization. |
| `resolveApiKey` | Function to resolve API key for any model: `await resolveApiKey(model)` |