P2.2: Document write path implementation verification

Verified the complete write path implementation covering:
- POST /indexes/{uid}/documents - add documents
- PUT /indexes/{uid}/documents - replace documents
- DELETE /indexes/{uid}/documents/{id} - delete by ID
- DELETE /indexes/{uid}/documents - delete by IDs array or filter

Key features verified:
1. Primary key extraction on hot path with 400 rejection
2. _miroir_shard injection before forwarding to nodes
3. Reserved field rejection (_miroir_shard always reserved)
4. Two-rule quorum (per-group quorum + degraded header)
5. Per-batch grouping for efficient fan-out
6. Independent shard routing for delete by IDs
7. Broadcast for delete by filter

All 34 tests pass (16 acceptance + 18 unit tests).

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

notes/miroir-9dj.2.md

@@ -0,0 +1,72 @@
+# P2.2 Write Path Implementation Summary
+
+## Overview
+The write path for Miroir is fully implemented, covering all four document operations:
+- `POST /indexes/{uid}/documents` - Add documents
+- `PUT /indexes/{uid}/documents` - Replace documents
+- `DELETE /indexes/{uid}/documents/{id}` - Delete single document by ID
+- `DELETE /indexes/{uid}/documents` - Delete by IDs array or filter
+
+## Implementation Location
+- **Route handlers**: `crates/miroir-proxy/src/routes/documents.rs`
+- **HTTP client**: `crates/miroir-proxy/src/client.rs`
+- **Core types**: `crates/miroir-core/src/scatter.rs`
+- **Routing logic**: `crates/miroir-core/src/router.rs`
+
+## Key Features Implemented
+
+### 1. Primary Key Extraction (Plan §3)
+- Extracts primary key from first document if not provided via query param
+- Tries common field names in order: `id`, `pk`, `key`, `_id`
+- Rejects batches without resolvable primary key with 400 `miroir_primary_key_required`
+
+### 2. `_miroir_shard` Injection (Plan §2)
+- Every document gets `_miroir_shard: shard_id` added before forwarding
+- Shard ID computed using `shard_for_key(pk_value, shard_count)` via rendezvous hashing
+- Field is stored as filterable attribute (set at index creation)
+- Stripped from all API responses
+
+### 3. Reserved Field Rejection (Plan §2, §5)
+- `_miroir_shard`: ALWAYS reserved (400 `miroir_reserved_field`)
+- `_miroir_updated_at`: Reserved when `anti_entropy.enabled: true`
+- `_miroir_expires_at`: Reserved when `ttl.enabled: true`
+- Non-reserved `_miroir_*` fields pass through
+
+### 4. Two-Rule Quorum (Plan §2)
+- Per-group quorum = `floor(RF/2) + 1` ACKs from that group's RF nodes
+- Write success if ≥ 1 group met its per-group quorum
+- `X-Miroir-Degraded: groups=N` header if ANY group missed quorum
+- HTTP 503 `miroir_no_quorum` only if NO group met quorum
+
+### 5. Per-Batch Grouping (Plan §3)
+- Documents grouped by target shard before fan-out
+- Each node gets exactly one HTTP request containing all docs it owns
+- Minimizes HTTP fan-out count (critical at scale)
+
+### 6. Delete Operations
+- **Delete by ID**: Routes each ID to its shard independently
+- **Delete by IDs array**: Groups IDs by shard for independent per-shard routing
+- **Delete by filter**: Broadcasts to all nodes (cannot shard-route)
+
+### 7. Additional Features
+- Alias resolution with multi-target alias rejection
+- Session pinning (Plan §13.6)
+- Task registry integration (P2.5)
+- Migration-aware routing with dual-write support
+
+## Test Coverage
+- **Acceptance tests**: 16 tests in `tests/p2_2_write_path_acceptance.rs`
+- **Unit tests**: 18 tests in `routes/documents.rs`
+- All 34 tests pass
+
+## Files Modified
+No source code changes were needed - the implementation was already complete.
+
+## Acceptance Criteria Met
+✅ 1000 docs indexed via POST — every doc fetch-by-id returns the same doc
+✅ Docs distribute across all configured nodes (no node < 20%)
+✅ Batch with one missing primary key → 400 miroir_primary_key_required
+✅ Doc containing _miroir_shard → 400 miroir_reserved_field
+✅ RG=2, RF=1, 1 group down: write to 1 group succeeds with X-Miroir-Degraded: groups=1
+✅ RG=2, RF=1, both groups down: 503 miroir_no_quorum
+✅ DELETE by IDs array routes each ID to its shard independently