The lowball-and-keep-applying advice is right, but the framing this thread keeps missing is that your quit salary is a 2024 reference price. The 2026 market reference for the same job description is structurally lower for three reasons: post-2024 layoffs increased available supply, the AI productivity multiplier shifts the output-per-comp frontier downward, and the senior band itself has recalibrated. Anchoring on what you used to make is generating most of the helplessness. Reset the anchor and the offers stop reading as personal rejection.

There's a second piece worth naming directly. Your replies in this thread saying there isn't much to learn about Cursor or Claude Code are exactly the signal the lowballing recruiters are picking up on. The tools aren't hard to operate. The senior signal is knowing when each one fails. That includes when Cursor's autocomplete beats Claude Code's agent loop and when it doesn't, how agent supervision overhead actually shows up on a 50k-LOC codebase, what MCP wiring buys you for non-trivial workflows, and the difference between prompts, CLAUDE.md instructions, and skill files. Two weeks of deliberate practice with one personal project per tool, deliberately hitting their failure modes, closes this gap. The "I'm above this" posture closes the door before the interview starts.

For the company that lowballed you a week ago, that engagement isn't dead unless they explicitly closed it. They invested 6 rounds in you. Re-engage today with something like "Thanks for the offer. I've been thinking it over. I'd be prepared to accept at $X if we can structure the first 6 months with a comp review tied to specific deliverables." That re-opens negotiation, signals confidence, and gives the hiring manager who already advocated for you a face-saving path. Walking away in silence after a week is the actual loss.