{"version":"2.1.267","anchor":"mcp-tools-can-now-declare-a-per-tool-result-size-cap-via-me","canonical_anchor":"mcp-tools-can-now-declare-a-per-tool-result-size-cap-via-me","heading":"MCP tools can now declare a per-tool result-size cap via _meta","tier":"use","area":"MCP","scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.267\/e\/mcp-tools-can-now-declare-a-per-tool-result-size-cap-via-me","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.267","markdown":"### MCP tools can now declare a per-tool result-size cap via _meta\n\nMCP tools can now declare their own result-size cap and other behavior via metadata\n\n**What**\n\nAn MCP (Model Context Protocol, the standard Claude Code uses to connect external tools) tool's manifest can now set several behaviors through `_meta` fields:\n\n- `anthropic\/maxResultSizeChars` overrides the tool's default maximum result size, up to an internal ceiling\n\n- `anthropic\/requiresUserInteraction` forces a permission prompt before the tool runs\n\n- `anthropic\/alwaysLoad` forces the tool to always be loaded\n\n- `anthropic\/searchHint` (or `promptOverrides.search_hints`) supplies a searchable hint string for the tool\n\n**Why**\n\nThis gives MCP tool authors finer control over how their tools behave in Claude Code, such as allowing larger results when needed, requiring explicit confirmation, or improving discoverability, without relying only on Claude Code's built-in defaults.\n\n- Area: MCP\n- Names: `_meta`\n- Tier: Use it now\n- Useful: 3\/5\n- Signal: 2\/5"}