Skip to content

Chat: no citation markers, and a thread's steps said in words - #83

Merged
abhaypadmanabhan merged 3 commits into
OrgoAI:mainfrom
abhaypadmanabhan:dev/abhay-plain-words
Nov 1, 2026
Merged

abhaypadmanabhan merged 3 commits into
OrgoAI:mainfrom
abhaypadmanabhan:dev/abhay-plain-words

Conversation

@abhaypadmanabhan

Copy link
Copy Markdown
Contributor

Replies and a thread's steps showed text meant for the model or for developers: "□cite□turn0search3□" after a sentence, and step chips like "ran python3 - <<'PY' from urllib…" or "clicked (666, 115)". A client reads that as broken.

Citation markers (private-use characters around what the model cites) are dropped from bot text in chat, thread notes and the sidebar preview. A thread's folded step chip says what happened ("Looked things up online", "Worked on the files", "Clicked on the screen", "Read example.com"); opening it still shows every raw step. lib/plain-words.ts. The wording follows how Grok Bot labels steps: short plain verbs, never the raw command.

Proof

  • scripts/test-plain-words.mjs (new): 2 passed, including a code sample (cd .) left untouched.
  • Screen: the 10-30 Four Corners brochure thread: notes without markers, chips "Looked things up online · 14 steps", "Worked on the files", "Clicked on the screen"; a reply with a marker and its sidebar preview show only the words.
  • Before: Bops 0.0.35 on a Mac, as a teammate in a shared workspace (2026-10-31). The same thread showed the markers and raw steps.
  • Not changed: the computer panel's step captions (computer.tsx, mac-computer.tsx).
  • Screens: a production build (next build, next start) of origin/main plus this and the other six PRs from the same review, with a local, uncommitted guard so the server never signs in or answers a call; a headless browser played a made-up copy of a shared workspace as a teammate with Write sees it.
  • npx tsc --noEmit -p . and eslint on the touched files: clean.

Review: Reviewed with sol-loop: 2 rounds with gpt-6.1-sol on high. Sol verdict: approved. Round 1 found "ran npm test" said "Set up the app"; fixed in the second commit (setup and history labels only for those commands). A nit (strip markers only from bot text) declined: people don't type those characters.

secscan: VERDICT: PASS. The npm audit's 1 critical and 8 high are already on main (package-lock.json unchanged).

One of seven small PRs from a review of Bops as BrandLord's staff would use it in a shared workspace; each stands alone.

🤖 Generated with Claude Code

abhaypadmanabhan and others added 2 commits October 31, 2026 16:17
A model's citation markers (private-use characters around "cite") showed in replies, thread notes and the sidebar preview; they go. A thread's folded step chip said its raw step ("ran python3 - <<'PY' ...", "clicked (666, 115)"); it says what happened ("Looked things up online", "Clicked on the screen"), and the steps open as they were. lib/plain-words.ts, scripts/test-plain-words.mjs.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Review: "ran npm test" said "Set up the app" and "ran git commit" said "Looked at the code history". Those
labels are now for installs and for git log, show, diff and blame; any other command says "Ran a command".

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@abhaypadmanabhan
abhaypadmanabhan merged commit 452e7d4 into OrgoAI:main Nov 1, 2026
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant