Memory that stays on-device and follows every agent
Mind gives your agents local, encrypted memory with provenance on every fact - and it follows you across agents instead of living inside one tool's silo.
The chain that removes it
Every new session starts from zero. Agents re-derive your conventions, re-ask settled questions, and forget the decisions you had already made.
Write facts once
A decision recorded once follows you across Claude Code, Cursor, Codex and Copilot - one memory, not four.
Carry provenance
Every fact stores its source and a code hash; when the code changes, the fact flags itself STALE automatically.
Keep it portable
The whole knowledge network exports as a signed .surfilpkg - at handoff, at audit, or on your way out.
Why the memory stays yours
On-device by construction
Memory syncs as end-to-end ciphertext; your source never leaves the machine to make recall work.
Honest when it does not know
Ask your codebase and get cited answers, or an honest empty response - never a confident wrong one.
No lock-in
Because it exports clean, the memory is yours to leave with, not a dashboard number that you rent.
More in this view
Same spine, a neighbouring starting point.
Cut token spend
Surfil trims the context agents never needed, measures the cache premium you are paying, and names the writes that are never read back - then signs the measured saving.
See this path →Block risky calls
Guard evaluates every tool call on-device before execution - catching secrets and prompt injection pre-flight, and failing closed whenever a call is ambiguous.
See this path →Prove compliance
Surfil turns agent activity into offline-verifiable evidence: signed attestations, control-mapped bundles, and a tamper-evident trail your auditors check themselves.
See this path →Turn on the first stage today
The install is reversible byte-for-byte, and the first signed receipt arrives after one normal week of traffic.