Source Intelligence
Sweep 28 Aug 2026 · 00:00Z Build v2.1.250 478 read Stable v2.1.236 Latest v2.1.250 Next v2.1.250 Feeds RSS JSON llms.txt

DisclaimerUnofficial, and not affiliated with Anthropic. Nearly all of this is read straight out of what ships: npm bundles, captured prompts, published docs. Anthropic's own notes go in verbatim, marked as theirs. The rest is my reading, and every entry carries the strings behind it. If one looks wrong, vote it down and say why.

Reference / Feature flags

tengu_hover_rest

A Claude Code feature flag, read out of the shipped bundle. It has been in the build since v2.1.221, including the newest one read.

What it is

The new storage layer now touches nearly every part of the product but is never actually created.

Not Anthropic's. This is the line from the earliest changelog entry here that named it, v2.1.227.

Presence across releases

ReleasesReadingDeclared typeDefault
2.1.245 – 2.1.250 5 releases never mined
2.1.221 – 2.1.243 22 releases in the build bool:false

A row reading never mined is a release with no archived bundle behind it. Nothing was read out of it, so nothing here says whether the name was in it. A declared type or a default is only ever what that release's own bundle stated.

Gate readings

ReleaseResolvedThis accountAnonymousCompiled default
v2.1.248 Off in both readings off off on
v2.1.242 Off in both readings off off not a boolean we can read
v2.1.238 Off in both readings off off on
v2.1.236 Off in both readings off off on
v2.1.235 Off in both readings off off on
v2.1.234 Off in both readings off off on
v2.1.227 read on another version Off in both readings off off on

Two arms, because they disagree: this account is the one account this site reads and anonymous is a baseline with no identity attached. A reading says nothing about your account, and off cannot rule out a rollout both readings sit outside of. What these readings can and cannot show.

Changelog entries naming it

  1. v2.1.248An in-memory storage backend is bundled but off by default

    A full in-memory storage backend ships, off unless you set an environment variable.

    gate cited by this entry

  2. v2.1.248A memory-only config write mode exists, and nothing in this build turns it on

    Config can save to memory only, but nothing in this build enables that mode.

    gate cited by this entry

  3. v2.1.248Transcript store fence is wired but nothing in this build can trip it

    A latch can make a session stop writing its transcript, but nothing can trip it here.

    gate cited by this entry

  4. v2.1.242On-disk storage layer gains symlink and permission rules per data type

    Where Claude Code stores your data gains per-type symlink and file-permission rules, off by default.

    gate cited by this entry

  5. v2.1.238Plaintext credential store gained an in-process cache

    Plaintext credential reads could come from memory instead of the file, but only under off-by-default storage.

    gate cited by this entry

  6. v2.1.238A second plugin asset-cache cleaner exists, switched off in this build

    An alternative plugin cache cleaner exists, but you still get the old file-timestamp scan.

    gate cited by this entry

  7. v2.1.238CLAUDE_CODE_HOVER_REST gives a local override for the new credential and settings storage

    You can pin the new credential and settings storage backend yourself with one environment variable.

    gate cited by this entry

  8. v2.1.238The v5 storage choice reads the environment variable first

    An environment variable now takes priority over the server flag when picking the new storage backend.

    gate cited by this entry

  9. v2.1.236Global config writes now pass the storage handle explicitly

    Settings writes now name their storage, so config can follow whichever backend a session uses.

    gate cited by this entry

  10. v2.1.236Cheaper storage listings and bounded settings watchers

    Listing your sessions is cheaper now that per-key statistics are skipped.

    gate cited by this entry

  11. v2.1.236Version 5 session storage is decided once per process

    Whether the newer session storage is used is decided once at startup and cannot change mid-session.

    gate cited by this entry

  12. v2.1.235Terminal recordings can be kept in the storage backend instead of loose files

    Terminal recordings could live in a storage backend instead of loose files, but that backend is off.

    gate cited by this entry

  13. v2.1.235Plugin version cache gets a staged publish path, behind the storage flag

    Plugin version caching could stage and atomically move entries, but only with the off-by-default storage backend.

    gate cited by this entry

  14. v2.1.235Command queue became an instance, and queued commands can be passive

    Queued commands can be marked passive, and Esc now lets other handlers run when it didn't cancel anything.

    gate cited by this entry

  15. v2.1.235Remote settings can be primed from local storage instead of probing disk

    Managed settings could be read from cached storage instead of disk, but the backend is off.

    gate cited by this entry

  16. v2.1.235More subsystems routed through an alternative storage backend that this build never creates

    Compaction, history purge, session moves, recordings and skill reads all gained storage-backed twins that never run.

    gate cited by this entry

  17. v2.1.234Credentials and transcript tails can be read through the storage backend

    Credentials and transcript tails can be read from the new storage backend, but the flag is off.

    named in this entry, gate cited by this entry

  18. v2.1.234Deleting a session from a transcript has a storage-backend implementation

    Deleting a session from a transcript has a new backend routine with retries, disabled here.

    named in this entry, gate cited by this entry

  19. v2.1.234Next-generation storage is pinned to one decision per process

    The new storage layer's on/off decision is now made once per process, and the backend is still an empty stub.

    named in this entry, gate cited by this entry

  20. v2.1.234Precomputed session resolution on resume, gated off

    Resume can report which session it resolved against an expected id, but the flag is off.

    named in this entry, gate cited by this entry

  21. v2.1.234Global config file watching can move onto the storage backend

    Your global config can be watched through the storage backend instead of the file, falling back if it fails.

    gate cited by this entry

  22. v2.1.234Resume reads the pre-compaction sidecar ahead of time

    Resume can fetch pre-compaction state early to cut a wait, but the flag is off.

    named in this entry, gate cited by this entry

  23. v2.1.234Plugin cache link materialization rewritten

    Turning plugin symlinks into real files now survives interruption and refuses unsafe targets.

    gate cited by this entry

  24. v2.1.234Restarting Claude Code hands the conversation over explicitly, and reports failure

    Restarting via /update or /tui now carries your conversation over explicitly and says so if the restart fails.

    gate cited by this entry

  25. v2.1.234Storage backend writes are now pinned behind a flag and pass explicit file modes

    Saved data can route through a new storage backend with explicit file permissions, but the flag is off.

    named in this entry, gate cited by this entry

  26. v2.1.227A new storage backend is wired through nearly everything, but never built

    The new storage layer now touches nearly every part of the product but is never actually created.

    gate cited by this entry

  27. v2.1.227A new storage backend is wired through the product but returns nothing

    A new storage layer is plumbed everywhere but builds nothing, so files are still read directly.

    gate cited by this entry

Documentation pages

None of Anthropic's documentation pages this site has a copy of carries this name.

Names beside it

Every feature flag in the inventory starting tengu, up to twelve of them.

tengu_1p_event_batch_configNo description from anybody. never mined tengu_account_plugins_sync_enabledPlugins you enabled on claude.ai can be pulled to your machine, but only your account server can enable it. never mined tengu_account_skills_sync_enabledSyncing skills from your claude.ai account stays off, but its checks were reshaped for other synced resources too. 2.1.234 tengu_adopt_claimA stale session claim now keeps its entries, marked stale, instead of throwing them away. never mined tengu_adopt_linkAdopting an agent's transcript now hard-links the file, copying only when linking is impossible. never mined tengu_adopt_relinkOn startup, subagent transcripts stored as symlinks are converted to hard links or copies. never mined tengu_advisor_unknown_contentAdvisor output you couldn't see now shows an error or a clear "declined to advise" message. never mined tengu_agent_hooks_origin_untrustedAgents in untrusted folders can no longer define inline MCP servers; trust the project to allow it. never mined tengu_agent_list_attachNo description from anybody. 2.1.138 tengu_agent_proxy_download_queueDownload backlog on the agent proxy is now reported as a telemetry event. never mined tengu_agent_tool_selectedSDK result messages now include a summary of the subagents your session started. never mined tengu_agent_worktree_cwd_escape_blockedAgent commands now have a concrete worktree boundary, so escapes get blocked with a clearer log line. never mined

Read out of the published npm bundle release by release, and out of Anthropic's own documentation as this site captured it. Nothing on this page is a description anybody here wrote about what the feature flag does. All feature flags.