stack-discovery

Stack discovery determines your project's technology stack by scanning manifests, CI/IaC files, and existing instructions to produce an evidence-backed stack fingerprint. It prevents wrong assumptions at the start of an AI coding session by confirming unknown details.

Evidence-based stack fingerprinting reads manifests, CI/IaC, and runtime signals to infer frontend, backend, database, and infrastructure components. It cites specific evidence for each detected component to avoid incorrect auto-inference in unfamiliar repositories.

Stack discovery uses confirmation-driven unknown handling, stopping to request one question at a time when evidence is missing or conflicting. This prevents the AI from making incorrect auto-inferences about your project's technology stack.

Yes, stack discovery specializes in Chrome extension runtime detection by identifying MV2 versus MV3 using manifest_version and corroborating build evidence. It detects common bundlers and build tooling specific to Chrome extension development.

Stack discovery generates a requirements-spec-ready output containing status, evidence, confidence, nextAction, and blockers for safe handoff to requirements intake or genesis workflows. You simply request stack discovery and confirm any missing details it asks for.

Automatic project intelligence requires manifest-based component detection and cannot infer stacks without explicit evidence. It stops and asks for explicit user confirmation for unknowns, meaning it will not guess when encountering missing or conflicting technology signals.