[jphein]

Server-side rules for the 2026-10-22 security scan (private advisory GHSA-72r8-ccwp-xcwq). Client and seed fixes went in #19.

• F8/F9: listings.create needs source null/community, no sourceId/sourceSyncedAt, and aggregates in createListing's range. Owner edits freeze source/sourceId/sourceSyncedAt.
• F3–F6 server-side: kind must be one of the 12 known kinds on create and on any kind edit. This also covers old APKs that keep the unguarded glyph lookup.
• F7: owners edit or delete only community pins, which closes the seed-race residual from fix(security): untrusted kind lookups, root error boundary, opaque default handle, seed skips owned rows #19.
• F1: flags.create binds the author link to the caller.
• Owner edits obey the same aggregate bounds as reporters (from an automated review).

Staging A/B on a throwaway self-hosted app with the repo schema (fields relaxed to optional, links identical), using an extended copy of creek-watch perms_ab.py:

• main's rules: 11 mismatches, exactly the scan findings (positive control);
• this branch: GATE PASS 77/77;
• an allow-all ruleset flips the 5 inverse-link probes to ALLOW, while main and this branch deny them (the forward row's rule is the one checked);
• every run ended with leftovers=0.

Prod prerequisites (read-only):

• prod rules == main (key order only), backed up before the push;
• prod schema links flags.author and saves.user to profiles, so no phantom namespaces;
• all 85 prod species and all 47,731 listings use one of the 12 kinds.

Reviews:

• Oracle (spyglass-ba) @ d7eb4a7: MERGE, nits applied (comments).
• Cross-model (gpt-6.1-sol): MERGE-AFTER-FIXES. (1) Inverse-link takeover: probed, denied under these rules, with a positive control. (2)/(3) Wording on aggregates and null exceptions: documented in the comments. (4) double() on numeric strings: pre-existing in the aggregate branch and not a range bypass; left as a residual.

🤖 Generated with Claude Code

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[coderabbitai]

Note

Currently processing new changes in this PR. This may take a few minutes, please wait...

⚙️ Run configuration

• Source mode: Auto
• Repository: Unavailable
• Branch: Unavailable
• File: No configuration file from this repository applied
• Commit: Unavailable
• Layers: defaults
• Review profile: CHILL
• Plan: Advanced
• Run ID: 9a7a896c-6a08-40c5-a97e-03a6a5763d9a

📥 Commits

Reviewing files that changed from the base of the PR and between a04352e and d3ac9ea.

📒 Files selected for processing: 1

• instant.perms.ts

+-------------------------+
| Please feed the models. |
+-------------------------+
  \
   \     /)/)
        ( '')
        /  |
      o(___)

✨ Finishing Touches

📝 Generate docstrings

• Commit to this branch
• Create a new PR

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[jphein]

Self-merged by vale (techempower lane) under JP's standing yes (2026-09-28). CI green; typecheck/lint clean; staging A/B GATE PASS 77/77, with main as the positive control (11 = the scan findings) and allow-all rules as the inverse-link control; gitleaks clean. Oracle (spyglass-ba): MERGE. Cross-model (gpt-6.1-sol): MERGE-AFTER-FIXES; each point probed or documented (see body). Prod perms push follows via realm-cloud bin/instant-dash, with prod rules backed up first.