From afe086356b08d4b3d15a0e00ad70466937942c3f Mon Sep 17 00:00:00 2001 From: Mark Shust Date: Tue, 6 Oct 2026 11:17:13 -0400 Subject: [PATCH] docs: describe mcp allow_destructive in terms of the destructive command marker #524 made run_console_command read #[Command(destructive: true)] instead of inferring from a force flag; the config comment still described the old rule. Co-Authored-By: Claude Opus 5.5 Claude-Session: https://claude.ai/code/session_01KyjR4cJTfro54iFZ9CWkv1 --- packages/mcp/config/mcp.php | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/packages/mcp/config/mcp.php b/packages/mcp/config/mcp.php index adc92880a..af080f6d2 100644 --- a/packages/mcp/config/mcp.php +++ b/packages/mcp/config/mcp.php @@ -22,8 +22,8 @@ 'queue:status', 'queue:failed', ], - // Lets run_console_command run an allowed command whose definition declares a `force` flag (db:reset, - // db:rollback, db:seed...). Off by default: the agent can pass --force itself, so the flag is no guard. + // Lets run_console_command run an allowed command marked #[Command(destructive: true)] (db:reset, + // cache:clear, queue:clear...). Off by default: the agent can pass --force itself, so that is no guard. 'allow_destructive' => Env::bool('MCP_ALLOW_DESTRUCTIVE', false), ], 'database' => [