Coder/deepseek-harness
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docs: structure model experience fields

Browse Source
This commit is contained in:
Tianyi Cui
2026-07-19 18:08:42 +08:00
parent bcc920369c
commit a278374e8f
92 changed files with 1334 additions and 512 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
packages/compact/compact/README.md
View File

@@ -61,19 +61,31 @@ Subclass `CompactService`, implement `compactIfNeeded` and `compactRegion`, and
### Conversation history, when a backend is invoked
**What the model sees**: A successful implementation replaces an older surface range with one user-role summary checkpoint; the raw events stay logged but stop appearing in derived model messages. The seam itself performs no rewrite.
#### What the model sees
**Token effect**: Zero direct tokens from this interface. A backend trades many retained history tokens for one summary and leaves the recent tail unchanged.
A successful implementation replaces an older surface range with one user-role summary checkpoint; the raw events stay logged but stop appearing in derived model messages. The seam itself performs no rewrite.
**KV Cache effect**: A successful backend replacement invalidates reuse from the first shadowed history token; the seam itself does not alter a request.
#### Token effect
Zero direct tokens from this interface. A backend trades many retained history tokens for one summary and leaves the recent tail unchanged.
#### KV Cache effect
A successful backend replacement invalidates reuse from the first shadowed history token; the seam itself does not alter a request.
### Transcript supplied to a compaction consumer
**What the model sees**: `renderTranscript()` joins entries with one blank line and renders them exactly as `User: <content>`, `Assistant: <content>`, `Tool result (call <callId>): <content>`, `Tool error (call <callId>): <content>`, `[Context: <content>]`, or `[Steering: <content>]`. Non-text blocks render exactly as `[reasoning: <text>]`, `[tool-call: <name>(<arguments>)]`, `[tool-result: <content>]`, `[tool-result]`, or `[<block-type>]`.
#### What the model sees
**Token effect**: Data-dependent input tokens are paid only by the auxiliary model or consumer that requests this transcript; the conversation model does not receive a duplicate transcript.
`renderTranscript()` joins entries with one blank line and renders them exactly as `User: <content>`, `Assistant: <content>`, `Tool result (call <callId>): <content>`, `Tool error (call <callId>): <content>`, `[Context: <content>]`, or `[Steering: <content>]`. Non-text blocks render exactly as `[reasoning: <text>]`, `[tool-call: <name>(<arguments>)]`, `[tool-result: <content>]`, `[tool-result]`, or `[<block-type>]`.
**KV Cache effect**: No conversation-cache invalidation. A consumer's auxiliary request can reuse only the exact prefix produced by this rendering; changed or compacted entries invalidate reuse from their first difference.
#### Token effect
Data-dependent input tokens are paid only by the auxiliary model or consumer that requests this transcript; the conversation model does not receive a duplicate transcript.
#### KV Cache effect
No conversation-cache invalidation. A consumer's auxiliary request can reuse only the exact prefix produced by this rendering; changed or compacted entries invalidate reuse from their first difference.
## Known Limitations and Deferred Work