The credible interval is the [[Bayesian]] equivalent to a [[confidence interval]]. Simply find the corresponding bounds on the [[posterior]].
As with confidence intervals, you would select the quantiles (bounds) to give the shortest interval for a given credible interval (typically this means excluding the tails).
In [[R]] use the `qnorm` function or its equivalent for the distribution of the posterior.
```R
alpha = 0.95
q = 1 - alpha/2
CI_lower = qnorm(q, mu, sig)
CI_upper = qnorm(alpha + q, mu, sig)
```