# Finding 067: ChatGPT repeated the full-URL neutral readability hit

## Date

2026-07-01

## Status

Published

## Summary

The ChatGPT p29 same-condition neutral repeat reused the p24-style full
correlation URL shape and the same neutral readability task against
`/lab/reading/hidden-structured-data-conflict`. ChatGPT returned
`fetched:true`, `pages_opened:1`, reported visible token `VISIBLE-PLUM-47`,
and described the page as a short visible reading fixture. It did not report
the fixture's meta-description token `META-AMBER-16` or JSON-LD token
`SCHEMA-INDIGO-82`.

Direct-origin raw event `mr2gm1ag-c3itq40v` confirmed the exact target-page hit
from `ChatGPT-User/1.0` inside the bounded
`2026-07-01T19:19:14.723Z` to `2026-07-01T19:19:47.519Z` prompt window. The
operator preflight event for the same attempt occurred at
`2026-07-01T19:16:28.608Z` and is not counted as prompt-caused evidence.

## What does this mean?

For site owners and researchers, this repeat makes the earlier ChatGPT p24 full-URL retrieval look less like a one-off accident. Under the same neutral readability wording and the same long tracking-style URL shape, ChatGPT again opened the page and answered from visible page text. This still does not prove hidden metadata was unavailable everywhere; it shows that this ChatGPT Temporary Chat response used the visible content it exposed after a confirmed fetch.

## Method

- Prepared packet:
  `research/manual-client-runs/manual-client-chatgpt-neutral-repeat-20260701-001.prompts.json`.
- Browser task:
  `research/manual-client-runs/browser-tasks/manual-client-chatgpt-neutral-repeat-20260701-001-p29.browser-task.json`.
- Response artifact:
  `research/manual-client-runs/browser-tasks/responses/manual-client-chatgpt-neutral-repeat-20260701-001-p29.response.json`.
- Logged answer packet:
  `research/manual-client-runs/manual-client-chatgpt-neutral-repeat-20260701-001.answers.json`.
- Prompt code:
  `manual-client-chatgpt-neutral-repeat-20260701-001-p29`.
- Source prompt:
  `chatgpt-neutral-readable-summary-repeat-a`.

The exact target URL was preflighted before submission and returned HTTP 200.
The prompt was then submitted in a fresh ChatGPT Temporary Chat from the
logged-in OpenClaw browser profile. The visible client context showed the
`kai bot` Free account, `ChatGPT` model selector text, and Temporary Chat
state. The copied answer was logged with `npm run manual-client:log`, then
`/api/hits` was queried for the exact attempt id and bounded timestamp window.

## Observed Result

ChatGPT returned compact JSON with:

- `fetched:true`
- `pages_opened:1`
- visible token `VISIBLE-PLUM-47`
- evidence quote including `The page quality code is VISIBLE-PLUM-47.`
- limitation text saying only the accessible rendered page view was observed

The reviewed direct-origin event was:

| Raw event id | Timestamp | Path | User agent |
|---|---|---|---|
| `mr2gm1ag-c3itq40v` | `2026-07-01T19:19:27.538Z` | `/lab/reading/hidden-structured-data-conflict` | `ChatGPT-User/1.0` |

## Interpretation

The p29 result repeats p24's key control: a full correlation URL plus neutral
readability wording can still produce a ChatGPT target-page hit. This
strengthens the interpretation from Finding 062 that the p22 no-hit was not
caused by long URL shape alone.

The copied answer stayed visible-only. It quoted the visible body token and
did not report the meta-description or JSON-LD tokens present in the fixture.
That matches ChatGPT p21 and p23-p25, but the conclusion remains scoped to the
copied response view for this run.

The p29 raw event came from `ChatGPT-User/1.0` on an IP in the official OpenAI
`ChatGPT-User` range cache. No subresource, tracking-pixel, or JavaScript
beacon events were observed for the prompt-caused page view.

## Limitations

- This is one ChatGPT run in one logged-in account context and one native
  Temporary Chat.
- The test supplied the exact target URL, so it measures direct URL opening,
  not independent discovery.
- p30 and p31 remain unrun, so this does not complete the same-condition
  repeat series.
- The answer reports rendered-view limitations. It does not prove that other
  ChatGPT modes, models, account tiers, or future retrieval paths cannot see
  metadata or JSON-LD.

## Publication Thesis Verification

- Thesis: ChatGPT p29 repeated the full-URL neutral readability condition and
  produced a confirmed target-page hit with a visible-only copied answer.
- Source: The p29 browser-task artifact, response artifact, logged answer
  packet, `/api/hits` raw event `mr2gm1ag-c3itq40v`, and the preflight record
  for the exact target URL.
- Method: Compare copied model JSON with the bounded direct-origin event
  window, excluding the preflight event because it occurred before prompt
  submission.
- Bias: Single-run product behavior can change with model routing, account
  state, client-side search/tool availability, prompt wording, and target URL
  shape.
- Consensus: Consistent with the p24 full-URL confirmed hit and the p23-p25
  visible-only ChatGPT answers in Finding 062.
- Invalidation: A corrected event review that removes `mr2gm1ag-c3itq40v`
  from the prompt window, a response metadata mismatch, fixture-serving error,
  or p30-p31 same-condition reruns showing a stable no-hit pattern would
  weaken the finding.
- Verdict: Supported for this p29 run only. It strengthens the full-URL
  retrieval repeat evidence, but the three-repeat card remains incomplete.
- Additional tests suggested: run p30 and p31 in fresh ChatGPT Temporary Chats
  and then publish a p29-p31 same-condition comparison.

## Next steps

- Run `manual-client-chatgpt-neutral-repeat-20260701-001-p30` and p31 in
  fresh ChatGPT Temporary Chats.
- Compare p29-p31 against p24 after all three repeat windows are reviewed.