Anyone using AI tools to cross-check their SMC structure reads? 3 weeks in with one of them, here is what holds up and what doesn’t.

Curious what tools people here are running to validate their SMC structure work. I picked up the habit of cross-checking my own reads against an AI analyzer after marking the wrong leg on a daily range twice last month. Posting in case it is useful, and in case someone has a better workflow than mine.

The one I have been running on BTC and ETH for the past 3 weeks: https://investisseur2-0.com/en/smc-ai-analyzer. Free, no signup, you paste a chart screenshot.

What it does well: liquidity pool tagging (SSL, BSL, equal highs and lows), FVG detection on H4 and Daily, last BOS and CHoCH, and a non-binding bias derived from the structural read. The FVG auto-marking is where it saves me time. I tend to miss small displacement FVGs when I am tired.

What it still fails on: OTE premium and discount when the swing range is ambiguous (which leg do you measure?), order block versus breaker block in choppy zones, Wyckoff phase C nuance (spring versus failed spring), and anything narrative-aware. It does not know ETF flows or macro context.

Sanity check on today's BTC tape: it tagged the $73.8K to $74K SSL pool as next draw, FVG April 24 at $76.8K to $77.4K unfilled below, $584M long liquidation cluster sitting on the sweep target. Matches what I had marked manually. The structural read is solid.

Where it falls short is flow context. Six straight days of BTC ETF outflows, $1.26B gone, 2026 net inflows collapsed from the $2.44B April record to $536M. Pure SMC reads structure not flow. The tool stayed structure-bullish while the flow says institutions are reducing risk.

Anyone else running their charts through an AI analyzer? What are you using, and where have you seen yours break? Especially curious about anyone running the same framework across FX and crypto.