geom_bar color according to 2 vars

I was looking how to set colors in the geom_bar ggplot2 function, but havn't seen yet similar data here a repro:

rm(list=ls())
library(ggplot2)
set.seed(123)
data=data.frame(hit=sample(1:3,20,replace = TRUE),
                type=sample(c("A","B","C"),20,replace=TRUE))
ggplot(data,aes(hit,fill=type))+
  geom_bar(stat="count",position=position_dodge())

I need to have some gradient from A to C and different colors from 1 to 3.

I think the following works for you which uses transparency to differentiate between the types:

library(ggplot2)

set.seed(123)
data <- data.frame(hit = sample(1:3, 20, replace = TRUE),
                   type = sample(c("A", "B", "C"), 20, replace = TRUE))

ggplot(data) +
  geom_bar(aes(hit, alpha = type, fill = as.factor(hit)),
           stat = "count", position = position_dodge()) +
  scale_fill_manual(name = "Hit", 
                    values = c("darkblue", "darkgreen", 'darkred')) +
  scale_alpha_discrete(name = "Type", 
                       breaks = c("A", "B", "C"), 
                       range = c(0.3, 1))

I also would recommend to have the complete dataset when plotting:

library(dplyr)
library(tidyr)
library(ggplot2)

data %>% 
  summarize(count = n(), .by = c(hit, type)) %>% 
  complete(hit, type = unique(type), fill = list(count = 0)) %>% 
ggplot() +
  geom_bar(aes(hit, count, alpha = type, fill = as.factor(hit)),
           stat = "identity", position = position_dodge()) +
  scale_fill_manual(name = "Hit", 
                    values = c("darkblue", "darkgreen", 'darkred')) +
  scale_alpha_discrete(name = "Type", 
                       breaks = c("A", "B", "C"), 
                       range = c(0.3, 1))