Besides the tools themselves, check out the details of adapter trimming in colorspace excerpted from one of the README files. Hopefully this is helpful.
Note that our approach to adapter trimming of color-space small RNA sequencing reads differs from that taken by Marco and Giffiths-Jones. In our approach, we directly identify the 3' adapter and trim it before attempting to map; Marco and Griffiths-Jones in contrast use a ".. sequential trimming and mapping approach". My sense is that the Marco/Griffths-Jones approach might result in imprecision in defining the 3' ends of the small RNAs -- since the exact size of small RNA matters a lot in many organisms, this could be a problem (although we really should directly compare the approaches -- to do at a later time) ... hence our approach of directly looking for the adapter.
The other issue with color-space small RNA data is the problematic nature of the first nucleotide .. both Marco and Giffiths-Jones and my document discuss the details of this issue. Essentially, we have been directly translating our trimmed reads into DNA-space before mapping to avoid the 1st nt issue. This causes a reduction in mappable reads, but guarantees a mapping result where both the biological 5' and 3' nucleotides of the original small RNA are accurately noted.
The SOLiD 3' trimming tool has options to output trimmed reads both in color-space and to translate the trimmed reads into DNA-space at the user's discretion.