docs: add recall/completeness use case to multi-agent MCP description (#5604)

## Summary
- Revised `futuresearch_multi_agent` tool description to cover both use
cases:
- **(a) Completeness/recall** — finding all items matching criteria
where a single agent has poor coverage
  - **(b) Depth** — parallel investigation across distinct angles
- Calls out list generation as a canonical use case with concrete slice
examples (region, vertical, time period)
- Supersedes #5600

## Test plan
- [ ] `test_manifest_descriptions_match_python_docstrings` passes
- [ ] Verify description reads well in Claude tool listing

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Sourced from commit e65221509a7cdc3d962e90ff57f18800da4f5132

Author: jackwildman

futuresearch-mcp/src/futuresearch_mcp/tools.py

@@ -325,11 +325,15 @@ async def futuresearch_multi_agent(
 ) -> list[TextContent]:
     """Deep parallel research: deploys multiple agents on different angles, then synthesizes.
 
-    Use this instead of futuresearch_agent or futuresearch_single_agent when the
-    question is hard enough to benefit from parallel investigation — e.g. when
-    the answer depends on genuinely distinct information sources, perspectives,
-    or methodologies that a single agent might miss or weight unevenly. Works on
-    a single question or across many rows of a dataset.
+    Use this instead of futuresearch_single_agent when either (a) completeness
+    matters — you're trying to find all items matching a criteria (e.g. "all AI
+    startups in Europe") and a single agent will have poor recall, or (b) depth
+    matters — the answer benefits from parallel investigation across genuinely
+    distinct angles (sources, geographies, methodologies) that a single agent
+    would weight unevenly. For list-generation tasks especially, directing agents
+    at different slices (by region, by vertical, by time period) dramatically
+    improves coverage before synthesizing. Works on a single question or across
+    many rows of a dataset.
 
     Each input row is researched by several direction agents in parallel (3-6
     depending on effort_level), each exploring a different angle. Their findings