tech-spikelisted

# Tech Spike Run a focused, time-boxed exploration to validate assumptions and uncover risks before committing to full implementation. The goal is fast learning, not production code. ## Process ### 1. Set up workspace Before any exploration: 1. Check if `.tmp/` exists relative to the project root. If not, create it. 2. Check if `.tmp` is in `.gitignore`. If not, append it. 3. All spike work happens inside `.tmp/` — never pollute the main project tree with exploratory code. ### 2. Understand the spike scope Read what the user has provided. They might give you: - A feature idea or set of requirements to validate - Explicit assumptions they want tested - A technical question they need answered - A design they're unsure about If the user hasn't provided explicit assumptions to test, generate them. Think about: - **Integration assumptions** — "Library X can do Y", "API Z supports this use case" - **Performance assumptions** — "This approach scales to N items", "Latency stays under X ms" - **Compatibility assumptions** — "This works across environments/versions" - **Correctness assumptions** — "Edge case X is handled", "Data format Y is valid" - **Feasibility assumptions** — "This can be done without modifying Z" Focus on the **highest-risk, highest-impact** assumptions — the ones that, if wrong, would invalidate the approach entirely. Present 3-7 assumptions ranked by risk, and get user confirmation before proceeding. ### 3. Check tool availability Before designing the