vrr/opencode
2
0
Fork 0
mirror of https://github.com/anomalyco/opencode.git synced 2026-08-08 15:53:19 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 6

feat(core): increase retained compaction context (#40906)

Browse source
This commit is contained in:
Aiden Cline 2026-08-06 13:45:06 -05:00 • committed by GitHub
parent ec95b27308
commit 8864b01d0b
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 4 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

2
packages/core/src/session/compaction.ts
View file

@ -20,7 +20,7 @@ import type { Info } from "../model"
import { SessionUsage } from "./usage"
const DEFAULT_BUFFER = 20_000
const DEFAULT_KEEP_TOKENS = 8_000
const DEFAULT_KEEP_TOKENS = 15_000
const OUTPUT_TOKEN_MAX = 32_000
const TOOL_OUTPUT_MAX_CHARS = 2_000
const SUMMARY_TEMPLATE = `Output exactly the Markdown structure shown inside <template> and keep the section order unchanged. Do not include the <template> tags in your response.

4
packages/www/content/docs/(Configure)/compaction.mdx
View file

@ -83,7 +83,7 @@ Add `compaction` to any [OpenCode configuration file](/config):
"auto": true,
"prune": false,
"keep": {
"tokens": 8000
"tokens": 15000
},
"buffer": 20000
}
@ -94,7 +94,7 @@ Add `compaction` to any [OpenCode configuration file](/config):
| --- | ---: | --- |
| `auto` | `true` | Runs the preflight context-size check. It does not disable manual compaction or one-shot provider-overflow recovery. |
| `prune` | None | Accepted by the V2 schema, but currently has no runtime effect. V2 does not prune old tool outputs in place. |
| `keep.tokens` | `8000` | Approximate number of tokens from the newest serialized conversation context to retain beside the summary. |
| `keep.tokens` | `15000` | Approximate number of tokens from the newest serialized conversation context to retain beside the summary. |
| `buffer` | `20000` | Safety reserve below an explicit input limit. Without one, it is the minimum context reserve and the model output allowance wins when larger. |
`keep.tokens` and `buffer` accept non-negative integers. Larger `keep.tokens`

2
packages/www/content/docs/config.mdx
View file

@ -329,7 +329,7 @@ Control automatic context compaction and how much recent context it preserves.
"compaction": {
"auto": true,
"keep": {
"tokens": 8000
"tokens": 15000
},
"buffer": 20000
}