Skip to content
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 9 additions & 5 deletions selectors.json
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
{
"version": "0.3.5",
"lastVerified": "2026-03-11",
"version": "0.3.6",
"lastVerified": "2026-04-11",
"platforms": {
"chatgpt": {
"turns": [
Expand Down Expand Up @@ -34,7 +34,9 @@
"[data-testid='human-turn']",
"[class*='font-user-message']",
"[class*='human-turn']",
"[class*='HumanTurn']"
"[class*='HumanTurn']",
".font-user-message",
"[class*=\"user-message\"]"

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Wire new Claude selectors into runtime parsing

The selectors added here are not used by the actual turn parser, so this change does not fix Claude extraction in production. content.js still hardcodes Claude user/branch selectors in getClaudeUserTurns/navigation logic, and selectorConfig from selectors.json is only referenced by diagnostic probe paths. In the regression scenario this commit targets (new Claude DOM classes), user turns can still be missed and the extension falls back to generic alternating-role parsing, producing incorrect conversation trees.

Useful? React with 👍 / 👎.

],
"assistantTurn": [
"[data-testid='assistant-turn']",
Expand All @@ -51,12 +53,14 @@
"branchPrev": [
"[aria-label*='prev' i]",
"[aria-label*='previous' i]",
"[aria-label*='上一']"
"[aria-label*='上一']",
"[aria-label*=\"previous\"]"
],
"branchNext": [
"[aria-label*='next' i]",
"[aria-label*='later' i]",
"[aria-label*='下一']"
"[aria-label*='下一']",
"[aria-label*=\"next\"]"
],
"scrollHost": [
"main"
Expand Down
Loading