Source Intelligence
Reading a new release v2.1.251 Analysing changes · 2/5 Sorting the findings · 1/4 steps 470 findings $18.91 so far

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.

All of v2.1.105 Home All releases olderv2.1.104 v2.1.107newer
Claude Code v2.1.105

Memory Rating Survey [In Development]

Feature flag
tengu_dunwich_bell Off in both readings

The flag server returned off for the account this site reads and for the anonymous baseline. A reading of off cannot rule out a rollout these two readings sit outside of.

This account: off · anonymous baseline: off · compiled default in v2.1.105: no gate table built for this version

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 v2.1.105. It isn't a statement about your account. What a flag value here can and cannot tell you

What

A "Did this help? (optional)" prompt allowing users to rate recalled memories as good or bad.

Status

Stubbed — the gate function H_5() always returns !1 (false). The UI components (rating buttons with thumbs up/down hover states) are fully built but unreachable.

Details
  • Rating buttons component exists with [Good]/[Bad] options
  • Keyboard shortcuts (+ and - keys) are wired up for rating
  • The infrastructure includes both memory-specific and general feedback survey paths
  • General feedback survey is also gated behind tengu_dunwich_bell (default: false)
Evidence

Stubbed gate (search for "Did this help?") and gate function returning !1

Strings lifted out of the shipped bundle, so the claim above can be checked against them.

See this entry in the whole of v2.1.105 →