github/openclaw
1
1
Fork 0
mirror of https://github.com/openclaw/openclaw.git synced 2026-04-03 03:03:24 -04:00
Code Issues Packages Projects Releases Wiki Activity

fix(transcript): always drop orphaned OpenAI reasoning blocks

Browse Source
This commit is contained in:
sebslight
2026-02-16 20:20:28 -05:00
parent 6229814af2
commit 83b1ae895e
2 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
docs/reference/transcript-hygiene.md
View File

@@ -95,7 +95,7 @@ external end-user instructions.
**OpenAI / OpenAI Codex**
- Image sanitization only.
- On model switch into OpenAI Responses/Codex, drop orphaned reasoning signatures (standalone reasoning items without a following content block).
- Drop orphaned reasoning signatures (standalone reasoning items without a following content block) for OpenAI Responses/Codex transcripts.
- No tool call id sanitization.
- No tool result pairing repair.
- No turn validation or reordering.

2
src/agents/pi-embedded-runner.sanitize-session-history.test.ts
View File

@@ -225,7 +225,7 @@ describe("sanitizeSessionHistory", () => {
expect(result).toEqual([]);
});
it("downgrades openai reasoning only when the model changes", async () => {
it("downgrades orphaned openai reasoning when the model changes too", async () => {
const { sessionManager, messages, modelId } = makeSnapshotChangedOpenAIReasoningScenario();
const result = await sanitizeWithOpenAIResponses({