Dennis Sirois

• You're scoping or building an MCP server and want to pressure-test tool boundaries, schema design, and transport choice before you commit.
• Your Claude Code project has outgrown ad-hoc prompts and you want real architecture: scoped skills, subagents, hook-based guardrails, file-based memory.
• You're wiring Claude to a proprietary or desktop app (CDP, ZMQ, custom bridge) and the standard examples stop where your problem starts.
• You're comparing Anthropic Agent SDK vs. Claude Code vs. raw API and want an honest read from someone who's shipped in all three.

Outcome: a concrete next-step decision — schema, transport, scope, or staging — written down before the call ends.

• You're about to point an agent at production and want to design the safety surface first: kill switches, mutating-vs-read separation, gated arming, state guards.
• You need an audit trail for an autonomous agent that survives a compliance review.
• You're designing inter-agent communication — file contracts, supervisor → executor channels, escalation patterns — and want to skip the obvious failure modes.
• You're stuck on a loop that "almost works" but drifts, loops, or hallucinates state, and you want a forensic look at the run.

Outcome: a sketched safety architecture, named failure modes, and the 2–3 things to instrument before you run again.

• You're an IT leader at a utility, hospital, financial firm, or regulated SMB being pushed to deploy LLMs without blowing the controls you already have.
• You're drafting an internal AI policy and want a practitioner's read on what's enforceable vs. theater.
• You're evaluating a "Claude inside our environment" vendor pitch and want a quiet second opinion on data exposure, network architecture, and audit posture.
• You need to bridge an IT/OT divide where the OT side will not touch the public internet — air-gapped, local-LLM, and proxy patterns.

Outcome: a scoped path forward — what's safe to pilot in 90 days, what to defer, and the 3 controls you shouldn't skip.

• Your team uses Claude Code for real work and keeps losing context between sessions — decisions get re-litigated, constraints forgotten, work repeated.
• You want a framework for maintaining institutional knowledge across session resets and team handoffs.
• You need a durable memory and continuity design before you scale a multi-session or multi-agent workflow.

Outcome: a session continuity design — what to capture, where to store it, how to keep it consistent — that fits your existing project structure.

• You're a prop trader, quant, or fintech founder weighing LLM-driven trading or research workflows and want a clear-eyed read on real vs. marketing.
• You're integrating Claude with a desktop trading or charting app (NinjaTrader, TradingView, MT5, Sierra) and need a working bridge pattern.
• You're scoping an RL pipeline (SAC, PPO, custom env, walk-forward) for a small team and want feedback on architecture and reward design first.

Outcome: a sharper picture of where AI helps, where it doesn't, and what the next 30 days should look like. (I'm not selling a strategy — the engineering is the point.)