meshbay.git/docs/third-review.md, branch 0.11

revert(hub): M6 — add_group_member must keep accepting node tokens

2026-09-01T18:50:45Z

M6 in the third review was a misread. `add_group_member` accepting a node-scoped token is deliberate (commit 0443cf8): the node calls POST /v1/groups/{id}/members/{username} after a CLI `member invite` so the group shows up in the invitee's SPA, authenticating with a node-scoped token. `group.admin_id == caller` is the real guard. An older test (`test_node_scope_blocks_add_member`) asserted the opposite and had been left red on main; the M6 "fix" (commit 6b38704) satisfied that test by switching the dependency to `require_user_scope` — which made `ops.create_invite`'s hub-membership call 403. That exception is swallowed with a log.warning, so an invited user silently never lands in group_members and the group is invisible to them. Reported from live testing (CLI `member invite grenet`, grenet saw nothing). Dependency back to `get_current_user`. The stale test now asserts the intended behaviour: a node token may add a member to its own operator's group (201) but not to a group it does not own (403). Third-review M6 marked WITHDRAWN. Co-Authored-By: Claude Sonnet 5 Claude-Session: https://claude.ai/code/session_011pG75yGK3NthNfyjH74omG

docs: mark M4 and M5 fixed in the third security review

2026-09-01T18:05:39Z

M4: federation `source_hub` bound to the token signer, push capped, revocation prunes the peer's own directory entries, state-changing MHP tokens are single-use. M5: a middleware adds a CSP and the other protective headers to every response, matching the desktop client's policy for these files. Every finding in the review (H1, H2, M1-M6) is now fixed; the summary, findings table and action plan reflect that. Original finding texts kept for the record. Co-Authored-By: Claude Sonnet 5 Claude-Session: https://claude.ai/code/session_011pG75yGK3NthNfyjH74omG

docs: mark M3 fixed in the third security review

2026-09-01T17:11:20Z

Link-preview SSRF surface bounded: per-connection + node-wide rate limit, port allowlist, connect-address re-check, decompression-bomb guard. Summary, findings table and action plan updated; original M3 text kept for the record. Co-Authored-By: Claude Sonnet 5 Claude-Session: https://claude.ai/code/session_011pG75yGK3NthNfyjH74omG

docs: mark M2 fixed in the third security review

2026-09-01T16:59:04Z

QUIC chat/stream handlers brought to WebRTC parity, and the QUIC listener gated off by default. Executive summary, findings table and action plan updated; the original M2 finding text is kept for the record. Co-Authored-By: Claude Sonnet 5 Claude-Session: https://claude.ai/code/session_011pG75yGK3NthNfyjH74omG

docs: add third security review (2026-09-01)

2026-09-01T16:18:05Z

Code-level review focused on what changed since second-review.md: the unified handshake, device linking, account recovery, email verification, reCAPTCHA, the hub instance-policy store, MHP federation, the relay registry, chat link previews, and the node's loopback control API. The second review's critical/high list is confirmed closed. New findings H1, H2, M1 and M6 are fixed in the preceding commits and annotated as such; M2 (QUIC chat handlers regress NS6/H1/H6), M3 (link-preview SSRF), M4 (federation trust), M5 (no SPA CSP) and the L-list remain. Co-Authored-By: Claude Sonnet 5 Claude-Session: https://claude.ai/code/session_011pG75yGK3NthNfyjH74omG