[aterga]

Summary

The ChatGPT plugin "Internet Computer (ICP)" was approved on 2026-10-13 and is published. This records that outcome in the OpenAI submission doc the way #218 recorded Anthropic's, and logs in both docs the late-September OAuth disruption that most likely caused OpenAI's first rejection. It also carries the one piece of #210 that was not superseded by #218, so #210 can be closed.

Related issues

Follows #218. Supersedes the remaining content of #210. References #191 and #203 (the switch to CIMD-by-default behind the disruption) and #189 (the earlier fix for the same ChatGPT redirect error).

Changes

• docs/openai-directory-submission.md:
  • New Submission log from the mailbox record: submitted 2026-09-30 as "Internet Computer"; rejected 2026-10-01 at sign-in or OAuth; the 2026-10-02 reply asking for a retry; resubmitted 2026-10-12 as "Internet Computer (ICP)", after which the review team answered that a rejection is only retried by resubmitting through the portal; approved 2026-10-13 (asdk_app_6acc9f589e548191a4ffa5a30842ac2d) and published by the submitter (date not recorded); OpenAI's 2026-10-28 notice that its daily MCP scan cannot run until the server authorization is reconnected on the plugin's MCPs tab.
  • The log says where the evidence stops: OpenAI's email does not say where its attempt failed, so the link to the CIMD switch is marked as an inference; the settings URL is recorded as what the submitter used to confirm publication, not as a documented public listing; the 2026-10-05 reply to the reporter is quoted for what it said.
  • Intro says approved and published; the review bullet says what the process looked like in practice; blocker 2 marks the pre-submission position superseded and records the SSO test account that went in (username only, no password); sections 4 and 5 no longer read as pending.
  • Checklist closed out, with three open items: reconnect the authorization for the daily scan and confirm a scan ran, verify openWorldHint in the next scan, and record the public listing link once known.
• docs/anthropic-directory-submission.md: one log entry for the OAuth disruption (2026-09-25 to early October): the switch to CIMD-by-default broke dynamic client registration for ChatGPT, a user reported it on 2026-09-25, OpenAI's reviewer reported being unable to complete sign-in on 2026-10-01, and on 2026-10-05 the reporter was told the migration was complete. This is Log the OAuth disruption and the first reviewer feedback on the Anthropic directory submission #210's surviving entry, re-verified against the mail thread and hedged where the evidence stops.

Testing

Docs-only change; no Rust or dashboard code touched.

• cargo build --locked --all-targets — not applicable
• cargo test --locked --all-targets — not applicable
• cargo fmt --all / cargo clippy --all-targets — not applicable
• npm test --prefix monitoring/mcp-status — not applicable

Every date, id and quotation in the new log entries was taken from the submission emails and the support thread in the submitter's mailbox, and the text was fact-checked against them a second time before the final commit.

Checklist

• I have read the Contributing guidelines.
• Docs (README / comments) updated for any user-visible change.
• No secrets, credentials, or internal-only information are included. The test account's password is not in the doc; the reporter of the September issue is not named.

🤖 Generated with Claude Code

https://claude.ai/code/session_01PruVfYuaS7ZAgNBMTsuRP6

[Copilot]

🔵 Needs a closer look

Submission details, quotations, publication status, and incident chronology rely on unavailable private mailbox and portal records and require verification by a maintainer with access.

No open issues 🎉

What changed in this PR

Records the OpenAI plugin’s approval and publication, closes out its submission record, and documents the OAuth disruption in both directory submission histories.

Changes:

• Adds OpenAI’s submission timeline, review outcome, and daily-scan authorization failure.
• Updates submission guidance and checklists, retaining follow-ups for scan reconnection, tool annotations, and the public listing link.
• Adds the OAuth disruption to Anthropic’s log, distinguishing reported events from inferred causes and unknown impact.

File	Description
docs/​openai-directory-submission.md	Records approval, publication, review history, and remaining operational follow-ups.
docs/​anthropic-directory-submission.md	Records the OAuth disruption overlapping the submission period and the limits of available evidence.

_{Review effort: Balanced}