Skip to content

docs: prefix-based access control guide#2839

Draft
jwhartley wants to merge 1 commit intomasterfrom
docs/prefix-access-control
Draft

docs: prefix-based access control guide#2839
jwhartley wants to merge 1 commit intomasterfrom
docs/prefix-access-control

Conversation

@jwhartley
Copy link
Copy Markdown
Contributor

Summary

  • Adds a new Access Control sidebar section, moving the existing "Authorizing Users" page there
  • Adds a new guide: Prefix-based access control covering the four layers of sub-prefix isolation and three common scenarios

Context

Customers frequently ask how to restrict users and tasks to sub-prefixes (e.g. staging vs. prod, regional isolation). The existing auth docs explain capabilities conceptually but don't cover practical setup or the interaction between the four layers (user grants, role grants, storage mappings, data plane access).

This is raw content for @aeluce to take over — location, structure, and format are all up for grabs.

Scenarios covered

  1. Default — single prefix, no isolation (what every new org starts with)
  2. Environment isolation with cross-read — dev/prod where dev tasks can read prod collections
  3. Full isolation — EU/US with separate storage, data planes, and per-user admin scope

Notes

  • Technically reviewed against internal Slack threads with Johnny (confirmed accuracy of the four-layer model, grant additivity, longest-prefix-match for storage/data planes)
  • Data plane UI visibility bug (RLS → GraphQL migration) is mentioned in a note
  • Private link limitation is called out in the Limitations section

@jwhartley jwhartley linked an issue Apr 8, 2026 that may be closed by this pull request
@github-actions
Copy link
Copy Markdown

github-actions bot commented Apr 8, 2026

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.

docs: add practical guide for sub-prefix access control

2 participants