Today ReCall 0.8.0 arrives with a new idea at its core: your codebase should remember โ€” not just what it is now, but everything it has been, tried, and learned along the way.

We call it a development memory layer. Here's what changed and why it matters.

From timeline to memory

ReCall has always captured a complete, microsecond-precision timeline of every file change. That's a powerful safety net. But a timeline alone isn't memory โ€” memory means the ability to answer questions about the past: what did we try, what failed, what did the last agent do, what's the last known good state?

0.8.0 turns the timeline into memory you can actually use.

What's new in 0.8.0

1. A persistent memory with a lifecycle

Every memory now carries a status โ€” created โ†’ verified โ†’ deprecated โ†’ archived โ€” plus a confidence label (HIGH / MEDIUM / LOW). Deprecated and archived memories are excluded from context and search, so stale knowledge can never be mistaken for current truth.

2. Experiment memory (the "don't try that again" layer)

Every rollback, undo, or redo now automatically records an experiment memory: "this change was tried and reverted โ€” N files, M lines." When an agent asks "what do we know about X" or fetches context, it learns what was already tried and rejected, instead of repeating the mistake.

3. Ask ROW โ€” your timeline, in plain English

A fully offline natural-language interface (no LLM, no cloud, no keys) that answers real questions from your timeline, memories, and agent sessions:

  • "What changed today?"
  • "Why was this file modified?"
  • "What did the last agent try?"
  • "What's the last known good state?"

Each answer links back to the exact timeline event as evidence.

4. Provenance โ€” every memory has receipts

Every memory card has a "View evidence" button that jumps to the timeline event it came from. Memory without evidence isn't trustworthy; now it's all traceable.

Why "memory" matters more than ever

Here's the shift in a sentence: ReCall used to answer "what happened?" โ€” now it answers "what do we know?" That turns a safety net into a compounding asset. The longer you use it, the smarter your workspace becomes โ€” and the smarter the AI agents working in it become, because they start each session with real context instead of discovering everything from scratch.

This is what we mean by the memory flywheel: every session makes the next one better.

Private by design

All of this runs locally. Your timeline, your memories, your agent sessions โ€” they never leave your machine. The database is encrypted at rest (AES-256 via SQLCipher). No cloud, no telemetry, no training on your data.

Git remembers what you committed. ReCall remembers everything else.