{"version":"2.1.2","anchor":"permission-explainer-gradual-rollout","canonical_anchor":"permission-explainer-gradual-rollout","heading":"Permission Explainer [Gradual Rollout]","tier":null,"area":null,"scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.2\/e\/permission-explainer-gradual-rollout","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.2","markdown":"### Permission Explainer [Gradual Rollout]\n\n**What**\n\nA new feature that analyzes tool use requests and displays risk assessments (LOW\/MEDIUM\/HIGH) with explanations when Claude requests permissions.\n\n**Details**\n\n- Evaluates commands before execution and categorizes risk levels\n- LOW risk: Standard dev workflows (git operations, tests, installs, builds)\n- MEDIUM risk: Bulk deletions, complex shell pipelines, config modifications\n- HIGH risk: Dangerous or irreversible operations, system-level changes\n- Only displays warnings for MEDIUM and HIGH risk actions\n- Controlled by `tengu_permission_explainer` feature flag\n- Can be enabled via `PERMISSION_EXPLAINER_ENABLED=true` environment variable\n\n**Evidence**\n\n`Hu2()` at line 417436 (permission explainer generator, contains `\"tengu_permission_explainer\"`)"}