NCBM/plyngent
1
0
Fork 0
mirror of https://github.com/NCBM/plyngent.git synced 2026-07-23 05:55:16 +08:00
Code Issues Packages Projects Releases Wiki Activity

core/agent: calibrate soft-compact with last-request prompt_tokens

Browse Source 
After the first model call, scale char-based token estimates by
API/resolved prompt size so budget checks track near-real tokens.
/compact uses the same calibration when available.
This commit is contained in:
worldmozara
2026-07-15 11:11:15 +08:00
parent c9e805f38d
commit b1223c8189
6 changed files with 128 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files
CLAUDE.md
+1 -1
View File
@@ -54,7 +54,7 @@ Async SQLAlchemy + aiosqlite. `MemoryStore`: schema init (+ lightweight SQLite `
- **`ChatClient`** Protocol for `chat_completions`.
- **`@tool` / `ToolRegistry`**: decorator infers JSON Schema from type hints; execute tools by name.
- **`run_chat_loop`**: multi-round tool loop; default **streaming** text deltas + stream tool-call merge; parallel tools; tool-result char budget; soft context compact on request; cooperative cancel points; optional `on_limit`.
- **`run_chat_loop`**: multi-round tool loop; default **streaming** text deltas + stream tool-call merge; parallel tools; tool-result char budget; soft context compact on request (**API-calibrated** after first usage when available); cooperative cancel points; optional `on_limit`.
- **`ChatAgent`**: optional `MemoryStore` (persist on success only); `stream`; system prompt; `pending_retry_text` + `retry()`.
- **`/compact`**: soft-compact tool dumps → model summary (no tools) → **new** session seeded with summary message.
- Events: text_delta, assistant_message, tool_call/result, max_rounds, **error** (`retryable`/`source`), **cancelled** (`reason`), **usage** (`TokenUsage`).