A signal is evidence, not an instruction
A buy, sell, or hold label removes most of the reasoning that matters. A useful signal also identifies its timeframe, confidence, source, freshness, and the conditions that would weaken it.
Portfolio fit changes the answer
The same signal can be reasonable for one portfolio and inappropriate for another. Existing concentration, available liquidity, risk tolerance, and correlated positions should be reviewed before a signal becomes a proposal.
Rejection is useful information
When a risk control blocks a proposal, the reason should be visible. A rejected action can expose concentration, position sizing, stale data, or market conditions that deserve attention even when no trade occurs.