A stack is only real if it can be recovered. Storage consolidation began as an organisation problem and became a resilience problem: where are the projects, where are the secrets, where are the deploy scripts, and what survives if a machine does not?
Where it is now
The T-drive includes a recovery-bundle workflow that collects priority roots, sensitive hits, manifests, and included-root lists into a generated backup bundle. It is designed around the practical idea that a BitLocker-protected USB should be able to carry enough context to restore momentum.
The barriers
The hard part is deciding what matters without leaking what should stay protected. Keys, vaults, tokens, environment files, MCP configs, project roots, and deployment scripts all matter, but they need handling rules and manifests.
How I did it
The bundle script searches known roots and sensitive patterns, copies priority data, writes manifests, and records status. This is not glamorous work, but it is the kind of work that makes the rest of the portfolio credible. Resilience is a feature.