Instruct CC agent to prefer objective metrics when ranking (#4874)

Lately, the CC agent has been prone to ranking by subjective 0-100
scores, even when there's an objective metric. Two examples are

1. ranking AI startups by revenue growth
2. ranking AI analysts by AGI timelines

Instead of using revenue growth rate or predicted year for reaching AGI,
CC would produce an arbitrary score column ranging from 0 to 100. Now
both examples give me the expected output field.

I remember fixing this for Autocohort a few months ago. There might be
other instructions from the Autocohort prompts that are worth copying
into the system prompt or skill for CC. But I'll leave that for a future
PR.
Sourced from commit ee9f0c910c0be47b50605fb716eb12a107edac4a

Author: rgambee

everyrow-mcp/src/everyrow_mcp/tools.py

@@ -448,9 +448,9 @@ async def everyrow_rank(params: RankInput, ctx: EveryRowContext) -> list[TextCon
     criteria are qualitative.
 
     Examples:
-    - "Score this lead from 0 to 10 by likelihood to need data integration solutions"
-    - "Score this company out of 100 by AI/ML adoption maturity"
-    - "Score this candidate by fit for a senior engineering role, with 100 being the best"
+    - "Estimate this drug's peak annual sales in billions of dollars"
+    - "What is this country's 5-year GDP growth rate as a percentage?"
+    - "Score this candidate from 0 to 100 by fit for a senior engineering role"
 
     This function submits the task and returns immediately with a task_id and session_url.