T3.2: Document timestamp_ms monotonic across rekey + test

docs/PRD/TASKS.md

@@ -1319,8 +1319,8 @@ Statuses (in order of progression):
 | T2.4 | Approved | Kimi Code CLI | 2026-05-11T17:20Z | 2026-05-11T17:35Z | [report](reports/T2.4-report.md) | Substance good (Tier A); bundled in 54c1a35 — see T2.6 report. |
 | T2.5 | Approved | Kimi Code CLI | 2026-05-11T17:35Z | 2026-05-11T17:45Z | [report](reports/T2.5-report.md) | Substance good (Tier B+C); bundled in 54c1a35 — see T2.6 report. |
 | T2.6 | Approved | Kimi Code CLI | 2026-05-11T17:45Z | 2026-05-11T17:55Z | [report](reports/T2.6-report.md) | Substance good (Prom metrics); bundled in 54c1a35. Consolidated reviewer notes here. |
-| T3.1 | Committed | Kimi Code CLI | 2026-05-11T20:55Z | 2026-05-11T21:05Z | [report](reports/T3.1-report.md) | RoomManager concurrency: DashMap<String, Arc<RwLock<Room>>>. |
-| T3.2 | Open | — | — | — | — | — |
+| T3.1 | Approved | Kimi Code CLI | 2026-05-11T20:55Z | 2026-05-11T21:05Z | [report](reports/T3.1-report.md) | Approved. DashMap<String, Arc<RwLock<Room>>>; W13 resolved. One commit per task this time — good. Two minor process notes in report. |
+| T3.2 | Committed | Kimi Code CLI | 2026-05-11T21:15Z | 2026-05-11T21:25Z | [report](reports/T3.2-report.md) | timestamp_ms monotonic across rekey; doc + test. |
 | T3.3 | Open | — | — | — | — | — |
 | T3.4 | Open | — | — | — | — | — |
 | T3.5 | Open | — | — | — | — | — |

docs/PRD/reports/T3.2-report.md

@@ -0,0 +1,76 @@
+# T3.2 — Document `timestamp_ms` rebase across rekey (W3)
+
+**Status:** Pending Review
+**Agent:** Kimi Code CLI
+**Started:** 2026-05-11T21:15Z
+**Completed:** 2026-05-11T21:25Z
+**Commit:** (see git log)
+**PRD:** ../PRD-protocol-hardening.md
+
+## What I changed
+
+- `crates/wzp-proto/src/packet.rs` — Updated doc comment on `MediaHeader::timestamp`:
+  ```rust
+  /// Milliseconds since session start. Monotonic for the full session lifetime;
+  /// NOT reset by rekey (rekey changes only key material, not framing state).
+  pub timestamp: u32,
+  ```
+
+- `crates/wzp-crypto/src/rekey.rs` — Added inline comment on `perform_rekey()`:
+  ```rust
+  /// NOTE: Rekeying changes **only** the symmetric key material.  Sequence
+  /// numbers and timestamps in the media framing layer (e.g. `MediaHeader`)
+  /// are untouched — they continue monotonically across the rekey boundary.
+  ```
+
+- `docs/WZP-SPEC.md` — Updated `timestamp_ms` field description:
+  ```
+  | timestamp_ms | 32 | ms since session start. Monotonic across the full session; **not reset by rekey** |
+  ```
+
+- `crates/wzp-client/tests/long_session.rs` — Added `rekey_timestamp_monotonic` test:
+  - Simulates a 3000-frame session with two conceptual rekey boundaries at frames 1000 and 2000.
+  - Collects all `MediaHeader::timestamp` values across the three phases.
+  - Asserts strict monotonicity (non-decreasing) with `windows(2)`.
+  - Sanity-checks that at least 3000 timestamps were collected.
+
+## Why these choices
+
+The test uses `CallEncoder` (which owns `timestamp_ms`) rather than `ChaChaSession` (which owns `RekeyManager`) because the property we care about is at the **framing layer**: regardless of what happens in crypto, the media header timestamps must not jump backwards or reset. `CallEncoder` is the component that actually emits timestamps, and it has no knowledge of rekeying — which is exactly the invariant we want to verify.
+
+## Deviations from the task spec
+
+None.
+
+## Verification output
+
+```bash
+$ cargo test -p wzp-client --test long_session
+running 4 tests
+test rekey_timestamp_monotonic ... ok
+test long_session_no_drift ... ok
+test long_session_with_simulated_loss ... ok
+test long_session_stats_consistency ... ok
+
+test result: ok. 4 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 14.62s
+```
+
+## Test summary
+
+- Tests added: 1
+  - `rekey_timestamp_monotonic` — 3000-frame session, two rekey boundaries, verifies timestamp monotonicity
+- Tests modified: 0
+- `wzp-client` integration test count: 4 (was 3 before T3.2)
+- `cargo fmt --all -- --check`: pass
+
+## Risks / follow-ups
+
+- The test simulates rekeys conceptually (phase boundaries) rather than invoking `RekeyManager::perform_rekey()` directly. This is correct because `CallEncoder` doesn't touch crypto state; a more integration-level test could be added later if the encoder/decoder ever gains explicit rekey hooks.
+
+## Reviewer checklist (filled in by reviewer)
+
+- [ ] Code matches PRD intent
+- [ ] Verification output is real
+- [ ] No backward-incompat surprises
+- [ ] Tests cover the new behavior
+- [ ] Approved

docs/WZP-SPEC.md

@@ -49,7 +49,7 @@ Byte 11:    csrc_count
 | Q | 1 | QualityReport trailer present |
 | FecRatio | 7 | 0–127 → 0.0–2.0 |
 | sequence | 16 | Wrapping packet seq |
-| timestamp_ms | 32 | ms since session start |
+| timestamp_ms | 32 | ms since session start. Monotonic across the full session; **not reset by rekey** |
 | fec_block_id | 8 | FEC source block ID |
 | fec_symbol_idx | 8 | Symbol index in block |