One way to possibly include type annotations to Snakemake workflow definitions. One could for example use ontologies like edam, as well as mime types.

snakemake-types

filetypes:
  # this directive automatically generates flags for each type that can be used in the rules below
  fq="edam:format_1929"
  fa="edam:format_1930"
  bam="edam:format_2572"

rule bwa:
    input:
        fq("genome.fa"),
        fa("reads/{sample}.fq")
    output:
        bam("mapped/{sample}.bam")
    conda:
        "envs/mapping.yaml"
    shell:
        "bwa mem -f {input} | samtools view -Sb - > {output}"
        
rule sort:
    input:
        # Snakemake would display an error message if the rule creating this input file has not flagged it as bam.
        bam("mapped/{sample}.bam")
    output:
        bam("mapped/{sample}.sorted.bam")
    conda:
        "envs/mapping.yaml"
    shell:
        "samtools sort {input} > {output}"