Follow Discord
Sweep 06 Nov 2026 · 01:32Z Build v2.1.319 537 read Stable v2.1.311 Latest v2.1.318 Next v2.1.319 Feeds RSS JSON llms.txt llms-full.txt Unofficial

Claude Code v2.1.282 ·

Tool-input echo setting is fixed for each session and carried into resumed sessions

Whether tool inputs are repeated back to the API is now decided once per session, and resumed sessions keep that and other prompt-format choices

Group of 2 Under the hood Internal Changes
JSON All of v2.1.282
Under the hoodTier: how much it should matter to you
2Useful: my rating, 1 to 5
3Signal: worth watching, 1 to 5
SessionsArea: what it touches
Internal ChangesKind: in v2.1.282,
Internal ChangesSection of the release

Unclear What each of the saved layout choices changes in practice is not settled.

What

Claude Code has a setting, echoWireToolInputs, that decides whether the inputs of tool calls (the actions Claude takes, such as reading or editing a file) are repeated back to the API in the conversation history it sends. Before this release it was read fresh on every request, so a remote change could flip it in the middle of a session. Now:

  • echoWireToolInputs is decided once per session, from the remote setting tengu_humble_hammock or the environment variable CLAUDE_CODE_HUMBLE_HAMMOCK, and then saved.
  • A resumed session inherits the value recorded for it instead of reading the setting again.
  • The saved prompt snapshot, a record of how the session's prompt was put together, now also holds systemTurns, toolChangeHeader, inlineTools, keptReminders, echoWireToolInputs and contextRendering (either "announced" or "inline"). Before, it held only the system prompt, the tool list, cliPrefix and reminderFold. When a session resumes, these are read back so the conversation keeps its format.
  • Claude Code now logs whether a value was recorded, whether it was inherited, where it came from, and when a recorded value is replayed.

Saving the value also depends on a second remote switch, tengu_foamy_spring, which falls back to on when the server sends no value; nothing has been read about it yet. For tengu_humble_hammock, the previous release built it in as off by default, and the flag server has since returned it as on for this site's account and for an anonymous baseline account; no reading has been taken under this release yet.

Why

In practice, the conversation history Claude Code sends stays in one consistent form for the whole life of a session, including after you resume it, instead of changing shape when a remote setting changes partway through.

Read from
Feature flag
tengu_humble_hammock Off by default, switched on for this account

The shipped code defaults this off, and the flag server returned on for the one account this site reads on this version. That is the reading that makes the entry above worth a second look, and it still says nothing about your account.

This account: on · anonymous baseline: on · compiled default in v2.1.282: off

These values were read against a different version of Claude Code, so treat them as the nearest reading available instead of one taken on this release.

Read once, for one account on one subscription tier, against v2.1.282. It isn't a statement about your account. What a flag value here can and cannot tell you

What has happened since
Flag reading moved The flag server now returns on for tengu_humble_hammock, read as this account. A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.
Flag reading moved The flag server now returns off for tengu_humble_hammock, read as this account. A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.
Flag reading moved The flag server now returns on for tengu_humble_hammock, read as this account. A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.
Flag reading moved The flag server now returns off for tengu_humble_hammock, read as this account. A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.
Flag reading moved The flag server now returns on for tengu_humble_hammock, read as this account. A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.
Flag reading moved The flag server now returns off for tengu_humble_hammock, read as this account. A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.
Flag reading moved The flag server now returns on for tengu_humble_hammock, read as this account. A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.
Flag reading moved The flag server now returns off for tengu_humble_hammock, read as this account. A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.

See this across every release →

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 doubtWhat each of the saved layout choices changes in practice is not settled.

See this entry in the whole of v2.1.282 →

Feedback