Commit graph

2 commits

Author SHA1 Message Date
Ben
866cc988b5 fix(dashboard-auth): use fixed-length sig suffix in stub token framing
The stub auth provider's _sign/_unsign helpers joined payload and HMAC
with a 'b"."' separator and recovered the parts via bytes.rsplit. HMAC-SHA256
digests are random bytes, so ~12% of the time the digest contains 0x2E
('.') and rsplit picks the wrong split point -- HMAC verification then
spuriously rejects valid tokens.

test_stub_refresh_round_trips was failing ~25% of the time in isolation
because of this.

Switch to a fixed-length suffix (32 bytes, sliced off in _unsign): no
separator means no collision class. After the fix, 10/10 runs pass.
2026-05-27 02:12:27 -07:00
Ben
628a52fce2 test(dashboard-auth): stub auth provider for E2E gate testing
Phase 2, Task 2.1. Self-contained fake IDP — start_login redirects
straight back to {redirect_uri}?code=stub_code&state=<s> so tests can
walk the OAuth round trip in-process. Tokens are HMAC-signed JSON blobs
(not real JWTs) — enough structure for verify_session to detect tamper
and expiry without pulling in pyjwt.

Lives in tests/ only — never registered as a real plugin. Phase 3's
end-to-end tests import StubAuthProvider directly.

Convention: exp <= now counts as expired (TTL=0 means born-expired)
— matches what Phase 6's silent-refresh test will need.
2026-05-27 02:12:27 -07:00