Authenticated session handling

When it applies

The interesting bugs live after login — cross-tenant IDOR/BOLA, mass assignment, injection on the real data APIs — but the run keeps collapsing to pre-auth findings because login is gated by MFA/OTP, or the token expires and every automated request 401s halfway through. This skill is the setup step that makes authenticated testing actually work. It does not attack the auth mechanism (web-oauth, web-auth-jwt, api-auth-attacks do that) — it obtains one legitimate session and keeps it usable.

Why it works

Post-auth coverage silently disappears when tooling can't hold a session: the scanner logs out, the token lapses, or MFA can't be satisfied autonomously. Capturing the session as a reusable artifact — the cookie/localStorage state for browser-driven testing and the raw bearer token for direct API replay — decouples "log in once" from "test a thousand requests," so the whole authenticated surface stays reachable for the life of the token.

Method

1. Get a real test account, in writing. For a pentest, the client provides it; for bounty, use an account you own (register two: attacker + victim, for IDOR diffing). Record it by reference in scope.txt — never paste live credentials into notes or a prompt.
2. Handle the MFA/OTP gate by the least-privileged means available:
   • TOTP seed shared → generate the current code from the Base32 seed (oathtool --totp -b <SEED>) and submit it. Fully repeatable.
   • Operator-relayed OTP → drive login to the OTP prompt, a human supplies the code once, continue.
   • Out-of-band / human-in-the-loop → persist the pre-OTP browser context, pause, let the operator complete the challenge, then resume and capture state.
3. Capture the session as artifacts:
   • Browser state — Playwright storageState.json (context.storage_state(path=...)) holds cookies + localStorage.
   • API replay — the raw access/ID token plus its expires_at.
4. Feed it to your tools:
   • Burp/Caido — a session-handling rule + macro that re-auths on a logged-out signature (a redirect to /login, a 401, a specific body), so long scans don't drop the session.
   • Playwright — launch with storage_state="storageState.json".
   • API fuzzers (ffuf, api-fuzzing) — inject the bearer as a header.
5. Refresh, don't re-login. Record expires_at; when a refresh token exists, refresh; only re-run the full login when it doesn't. A wave of sudden 401s means the token lapsed — refresh before you trust "not vulnerable."
6. Diff two sessions for access control. Hold A's and B's artifacts side by side and replay each other's object references (web-idor, api-bola) — the core authenticated test.

Gotchas

• Present creds ≠ working creds. Before spawning a big authenticated scan, verify the session actually reaches a known post-auth endpoint (200 with your data) — otherwise you "test" a logged-out surface and report false negatives.
• Never inline the token. Pass sessions by file reference; keep secrets out of prompts, notes, and any committed file. The engagements/ tree is git-ignored — keep them there.
• If a realm truly can't be authenticated (no test account, no OTP relay), say so honestly — record the post-auth surface as deferred/untested, don't fake an N/A or silently skip it.
• Scope the session's blast radius — a test account with real admin rights can do real damage; prefer least-privilege test roles, and treat destructive actions per RoE.

Verify success

A request built entirely from your captured artifact (no live browser) returns a known post-auth response with your account's data — and a two-session IDOR probe runs cleanly against it.

References

Playwright authentication/storageState docs; Burp/Caido session-handling rules; RFC 6238 (TOTP).