Skip to main content
Glama
PROGRESSIVE_DISCLOSURE.md4.22 kB
# Progressive Disclosure Control response verbosity across all tools using the standardized `response_mode` parameter. ## Overview All tools support `response_mode` for token efficiency: | Value | Description | Token Range | Use Case | |-------|-------------|-------------|----------| | `minimal` | IDs, counts, status only | Smallest | Quick checks, validation | | `standard` | Structure + key details (default) | Moderate | Most operations | | `full` | Complete data and details | Largest | Analysis, reporting | **Token Savings**: Significant reduction using `minimal` vs `full` ## Standard Values ### `minimal` Returns only essential data: - **execute_query**: Row count + schema (no data) - **health_check**: Component status only - **get_catalog_summary**: Object counts only - **get_report**: Metadata (title, ID, counts) - **evolve_report**: Operation counts only ### `standard` (default) Balanced response for typical workflows: - **execute_query**: Sample rows + metrics - **health_check**: Status + remediation guidance - **get_catalog_summary**: Counts + database breakdown - **get_report**: Full outline with IDs - **evolve_report**: Counts + IDs of changed items ### `full` Complete data: - **execute_query**: All query results - **health_check**: Status + remediation + diagnostics - **get_catalog_summary**: All statistics + distributions - **get_report**: All content + audit trail - **evolve_report**: Complete change details ## Tool-Specific Modes ### execute_query | Mode | Description | Savings | |------|-------------|---------| | `schema_only` | Column schema only, no rows | Maximum | | `summary` | 5 sample rows + metrics | High | | `sample` | 10 sample rows | Moderate | | `full` | All query results | baseline | ### get_report Legacy mode values mapped to standard: - `summary` → `minimal` - `sections` / `insights` → `standard` - `full` → `full` ## Usage Examples ### Quick Checks (Default Behavior) ```python # Execute query - defaults to 'summary' (5 sample rows + metrics) execute_query( "SELECT * FROM customers", reason="Quick validation" ) # Returns: 5 sample rows + key_metrics + hint to use response_mode='full' # Check system health (defaults to minimal) health_check() # Verify catalog exists (defaults to minimal) get_catalog_summary() # Check report exists (defaults to minimal) get_report("Q1 Revenue") ``` ### Explicit Mode Control ```python # Get full report structure get_report("Q1 Revenue", response_mode="standard") # Get all query results (no sampling) execute_query( "SELECT * FROM table", reason="Full export", response_mode="full" ) ``` ### Complete Data ```python # Get all results (explicit) execute_query( "SELECT * FROM table", reason="Export", response_mode="full" ) # Get full report content with audit trail get_report("Q1 Revenue", response_mode="full") ``` ## Backward Compatibility Legacy parameter names are supported with deprecation warnings: | Tool | Legacy Parameter | Status | |------|-----------------|---------| | `execute_query` | `result_mode` | Deprecated | | `health_check` | `detail_level` | Deprecated | | `get_catalog_summary` | `mode` | Deprecated | | `evolve_report` | `response_detail` | Deprecated | | `evolve_report_batch` | `response_detail` | Deprecated | | `get_report` | `mode` | Deprecated | **Migration**: Replace legacy parameters with `response_mode`: ```python # Old (deprecated, shows warning) health_check(detail_level="minimal") # New (recommended) health_check(response_mode="minimal") ``` ## Best Practices 1. **Start minimal, drill down**: Check existence first, then fetch details if needed 2. **Use standard for most operations**: The default is optimized for typical workflows 3. **Reserve full for exports**: Only request complete data when necessary 4. **Combine with filters**: Use `min_importance`, `section_ids`, etc. with `response_mode` ## See Also - [Tool Index](./TOOLS_INDEX.md) - Complete tool reference - [execute_query](./tools/execute_query.md) - Query execution with result modes - [get_report](./tools/get_report.md) - Report retrieval with progressive disclosure - [Error Handling](./ERROR_HANDLING.md) - Error handling patterns

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/Evan-Kim2028/igloo-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server