Follow Discord
Sweep 22 Oct 2026 · 04:32Z Build v2.1.308 527 read Stable v2.1.308 Latest v2.1.308 Next v2.1.308 Feeds RSS JSON llms.txt llms-full.txt Unofficial

Claude Code v2.1.304 ·

Memory extraction can keep its place across worker restarts

A new carry_across_restarts option lets post-turn memory extraction save and restore its progress when its worker restarts

Group of 2 Use it now Notable No documentation found New Features
JSON All of v2.1.304
Use it nowTier: how much it should matter to you
4Useful: my rating, 1 to 5
4Signal: worth watching, 1 to 5
MemoryArea: what it touches
New FeaturesKind: in v2.1.304,
What probably matters to youSection of the release

Unclear It is not known exactly what is kept across a restart when carry_across_restarts is true.

What

Claude Code can pull memories out of a conversation after each turn. The settings for this, read from CLAUDE_CODE_POST_TURN_MEMORY_CONFIG or a hosted config, gain a yes/no option, carry_across_restarts, exposed as carryAcrossRestarts. It is false when absent, and forced off when CLAUDE_CODE_POST_TURN_MEMORY_SYNC is set.

When it is on, the extractor that works through an MCP server saves and restores where it was:

  • its cursor, meaning how far through the conversation it has read
  • its turn count
  • the time of its next scheduled pass

A stored record holds turns, cursor, tail and passAt. A resumed worker logs "[extractMemories] (mcp) resumed from the previous worker:" with how many turns have ended since the last pass and whether the cursor was restored. Before, this was kept only in memory and lost on a restart.

Why

In practice, a memory worker that restarts can pick up where the last one stopped instead of starting from scratch. Nothing changes unless the config turns the option on.

Read from
Names in the bundleCLAUDE_CODE_POST_TURN_MEMORY_CONFIGcarry_across_restarts
How sure we are
Something disagreesSomething we can check disagrees with this entry, or the writer said they could not settle it.
The writer flagged doubtIt is not known exactly what is kept across a restart when `carry_across_restarts` is true.

See this entry in the whole of v2.1.304 →

Feedback