Fix MongoDB webhook workflow e2e test

[dustintownsend]

• Make hook creation atomic by using the unique index on token instead of a check-then-insert pattern, which had a TOCTOU race condition
• Properly catch MongoDB duplicate key errors (code 11000) and convert to 409
• Add debug logging to hooks create, get, and getByToken operations
• Add comprehensive webhook flow tests to the testing package:
  • Test creating and retrieving hooks with random tokens (like e2e test)
  • Test concurrent hook creation and lookup
  • Test that getByToken returns 404 for non-existent tokens
  • Test immediate visibility of hooks after creation

These changes help diagnose and potentially fix the MongoDB e2e test failure where webhookWorkflow gets 404 instead of 202.

[github-actions]

Benchmark Comparison

Cross-comparison of workflow performance across World implementations.

📈 Comparing against baseline from main branch. Green 🟢 = faster, Red 🔺 = slower.

workflow with no steps

World	Workflow Time	Wall Time	Overhead	vs Fastest
🥇 💾 Starter (in-memory)	0.048s (+4.4%)	1.009s (~)	0.961s	1.00x
🔴 Redis	0.053s (+28.4% 🔺)	1.009s (~)	0.956s	1.10x
🍃 MongoDB	0.081s (-24.1% 🟢)	1.009s (~)	0.928s	1.69x
💻 Local	0.086s (-5.2% 🟢)	1.009s (~)	0.923s	1.80x
🗄️ Turso	0.096s (+10.9% 🔺)	1.010s (-0.6%)	0.915s	2.00x

workflow with 1 step

World	Workflow Time	Wall Time	Overhead	vs Fastest
🥇 💾 Starter (in-memory)	0.138s (+1.3%)	1.007s (~)	0.870s	1.00x
🔴 Redis	0.154s (+29.7% 🔺)	1.007s (~)	0.853s	1.12x
💻 Local	0.259s (-6.0% 🟢)	1.008s (~)	0.749s	1.88x
🍃 MongoDB	0.289s (-3.9%)	1.007s (~)	0.718s	2.10x
🗄️ Turso	0.333s (+8.0% 🔺)	1.026s (+1.7%)	0.693s	2.42x

workflow with 10 sequential steps

World	Workflow Time	Wall Time	Overhead	vs Fastest
🥇 💾 Starter (in-memory)	1.016s (+1.3%)	1.607s (+14.2% 🔺)	0.590s	1.00x
🔴 Redis	1.120s (+32.7% 🔺)	2.007s (+99.5% 🔺)	0.887s	1.10x
💻 Local	1.903s (-3.8%)	2.007s (-9.2% 🟢)	0.104s	1.87x
🍃 MongoDB	2.145s (-0.8%)	3.011s (~)	0.866s	2.11x
🗄️ Turso	2.152s (~)	3.011s (~)	0.859s	2.12x

workflow with 10 parallel steps

World	Workflow Time	Wall Time	Overhead	vs Fastest
🥇 💾 Starter (in-memory)	0.334s (+1.5%)	1.007s (~)	0.673s	1.00x
🔴 Redis	0.405s (+44.2% 🔺)	1.006s (~)	0.601s	1.21x
💻 Local	0.640s (-5.9% 🟢)	1.007s (~)	0.367s	1.92x
🍃 MongoDB	1.577s (-3.3%)	2.007s (~)	0.429s	4.73x
🗄️ Turso	1.582s (~)	2.010s (~)	0.428s	4.74x

---

Stream Benchmarks

Stream benchmarks include Time to First Byte (TTFB) metrics.

workflow with stream

World	Workflow Time	TTFB	Wall Time	Overhead	vs Fastest
🥇 💾 Starter (in-memory)	0.230s (+2.1%)	1.006s (~)	1.008s (~)	0.778s	1.00x
🔴 Redis	0.253s (+31.2% 🔺)	1.003s (~)	1.007s (~)	0.754s	1.10x
🗄️ Turso	0.491s (+1.2%)	0.974s (~)	1.009s (~)	0.518s	2.13x
🍃 MongoDB	0.529s (+2.4%)	0.929s (-2.0%)	1.008s (~)	0.478s	2.30x
💻 Local	⚠️ missing	-	-	-	-

---

Summary: Average Performance by World

World	Avg Workflow Time	Benchmarks
💾 Starter (in-memory)	0.353s	5
🔴 Redis	0.397s	5
💻 Local	0.722s	4
🍃 MongoDB	0.924s	5
🗄️ Turso	0.931s	5

Column Definitions

• Workflow Time: Runtime reported by workflow (completedAt - createdAt) - primary metric
• TTFB: Time to First Byte - time from workflow start until first stream byte received (stream benchmarks only)
• Wall Time: Total testbench time (trigger workflow + poll for result)
• Overhead: Testbench overhead (Wall Time - Workflow Time)
• vs Fastest: How much slower compared to the fastest World for this benchmark

Worlds:

• 💻 Local: Built-in workflow world
• 🐘 Postgres: PostgreSQL database world
• ▲ Vercel: Vercel production world
• 💾 Starter: In-memory reference implementation
• 🍃 MongoDB: MongoDB database world
• 🔴 Redis: Redis/BullMQ backend