{"version":"2.1.290","anchor":"unknown-tool-input-keys-are-set-aside-instead-of-failing-val","canonical_anchor":"unknown-tool-input-keys-are-set-aside-instead-of-failing-val","heading":"Extra unknown details in a tool request are now dropped instead of failing","tier":"notice","area":"Tools","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.290\/e\/unknown-tool-input-keys-are-set-aside-instead-of-failing-val","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.290","markdown":"### Extra unknown details in a tool request are now dropped instead of failing\n\nWhen a tool request contains details the tool does not recognise, Claude Code can now drop them and try again\n\n**Unclear.** Whether anything other than the plan mode check uses this is not clear.\n\n**What**\n\nWhen Claude uses a tool, it sends the tool a set of details. Claude Code checks those details against what the tool expects. If the only problem is extra details the tool does not recognise, Claude Code now removes them and checks again. This happens only when the tool's own check passes, and only when a remote switch allows it; that switch is on unless it is switched off remotely.\n\nThis is used when checking whether a tool call is read-only in plan mode, the mode where Claude plans changes without making them.\n\n**Why**\n\nTool calls with stray extra details may now go ahead instead of producing an error.\n\n- Flag `tengu_tidy_eclipse`: Not enough to say (read for one account on one subscription tier against v2.1.290; this account: no value returned, anonymous baseline: no value returned, compiled default: 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.\n- Area: Tools\n- Tier: You'll notice\n- Useful: 2\/5\n- Signal: 2\/5\n- Scope: individual\n- Heads-up: no"}