Claude Code decides, for each tool Claude wants to use, whether to allow it, and sometimes asks you. It now sends a usage report, called telemetry, about each of those decisions. The report says what kind of tool it was, how many times the sandbox was broken (the sandbox is a walled-off space that limits what commands can reach) and what made the decision:
- a safety check
- a managed policy
- a permission rule
- analysis of the command
- a hook, a script you set up to run at certain moments
- auto mode blocking it
- auto mode giving no verdict
- the permission mode
- a sandbox override
It also reports when a hook pushes back after a tool has run. A remote switch, tengu_snoopy_creek, appears to control this, and it falls back to on when the server sends nothing.
This is a new stream of usage data sent on every permission decision.
tengu_snoopy_creek Not enough to sayNothing here resolved what this flag was doing on this version, so nothing here should be read as on or off.
This account: no value returned · anonymous baseline: no value returned · compiled default in v2.1.307: on
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 v2.1.307. It isn't a statement about your account. What a flag value here can and cannot tell you
New in this build: tengu_snoopy_creek