I recently hosted a discussion with special guest Mary Beth Steven on SWI and Assessment on my free weekly "SWI Digital Drop In". 
It was clearly a topic of interest to a large group of folks and is one deserving more attention. In a follow up discussion on that topic, I posted some thoughts on just one key aspect of assessment in SWI I call "spelling-out-orthography". You can see that discussion here if you are a member of Mary Beth's FB group "Structured Word Inquiry in the Classroom" or if you join it. 
Another member posted this great question to my post:
"How would you spell out the word <climb>? Would it be c-l-i-mb or c-l-i-m (b)as a silent letter?"
Other members of that FB group already offerd excellent responses, but this question prompted me to write a longer response that I decided to post in this publick forum so people can consider it whether or not they are on FB. 
The first thing that I love about this question is how it uses notation for "spelling-out-orthography" that makes it so clear what her question is. And her question is itself a good asssessment of current strong understanding of the questioner! 
The only thing I would suggest to add precision to this question would be to use the terms 'grapheme' and orthographic marker. If I wanted to re-write this question with more precision, I could pose it this way:
How would you spell out the base <climb>? Would it be "c-l-i-mb" with an <mb> digraph for the final /m/ phoneme, or "c-l-i-m-b" with the <m> as a single-letter grapheme  for /m/ and the <b> as an orthograhic marker?

But even without those little revisions, this question  helps us see how this practice provokes close scrutiny of orthography. We see it as a tool for constant practice of grapheme-phoneme correspondences as well as a means to highlight questions which we use to refine our on-going understanding of how our written word works.
Note that before we can announce the structure of this word, we need to recognize that it is a simple construction -- just a base. The conventions I recommend (linked here) state that in the base, we do our best to announce every orthographic structure. The only orthographic structures we will find in any base are graphemes and orthographic markers.
With that recognition, we need to do phonemic awareness to identify the phonemes of the base in order to associate them with the graphemes that we actually announce.
Note that we can only safely draw conclusions about graphemes when we are confident about the morphological structure. This is a great reminder about an inherent fact about graphemes -- the are always WITHIN morphemes.
The IPA of word "climb" for American English from my Oxford is written like this /klaɪm/. I can isolate those phonemes as /k/ /l/ /aɪ/ /m/.
The first three phonemes are straightforward to map to their graphemes in this word. So the initial part of our spelling out is "c-l-i...".
But the questioner is wondering how to deal with the orthography of "climb" and that final /m/ phoneme in this word.
I can think of two plausible hypotheses of how to understand the presence of this letter <b>.
One possibility is that it is part of a digraph <mb> for /m/. Searching the <mb> letter sequence in the Word Searcher shows us words like "lamb" "crumb" "limb" and many others. On first glance I notice that the letter sequence is always final in a morpheme when associated with the /m/ phoneme. So if <mb> is a digraph, I know to see if this is a grapheme that is strictly used finally for the /m/ phoneme, or perhaps I will find some counter examples.
Another possibility is that we have a single-letter grapheme <m> representing the phoneme /m/. That would make sense only if the letter <b> is not a grapheme (or part of a digraph <mb>), but instead an orthographic structure that can be called an "orthographic marker". Orthographic markers are not graphemes, but letters or combinations of letters that do not in and of themselves represent a phoneme in any of the members of a morphological family. Instead they provide some other meaningful function. In the case of "climb", it is a reasonable hypothesis that the <b> is not a grapheme or part of a grapheme, but an orthographic marker signalling a connection between "climb" and "clamber" as already mentioned. (Also consider limb/limber, bomb/bombard.) Personally, I can't say that I've come to a strong conclusion regarding the two plausible hypotheses of the orthography of this word.
Hypothesis 1) The "climb" has a single-letter grapheme <m> for the /m/ phoneme and the <b> is an orthographic marker linking to "clamber".
Hypothesis 2) "climb" has a digraph <mb> which is etymologically driven to provide a signal of the connection to "clamber".
To reflect hypothesis #1, we would spell-out "climb" as "c-l-i-m-b" with each structure (grapheme and orthographic marker) announced on its own.
To reflect hypothesis #2, we would spell-out "climb" as "c-l-i-mb" with each structure (only graphemes in this case) announced on its own.
Note that the questioner already reflected these hypotheses in her brief question! All I have done is add some precision to the linguistic language. 
It may well be that someone can provide clear evidence for which of these analysis is best. But until I understand how to do this, I am happy to discuss these two possibilities. And the process of that discussion will make it clear that there is good reason for that letter <b> in this word whatever that good reason is! 
But notice that if teachers put themselves in the position of "spelling-out the orthography" of words, they can't just ignore these little surprises. Spelling-out orthographic structure helps us recognize what we don't yet know. And we won't always have a 'clean answer' but we DO NOT say that's an exception to be memorized. Even without the full understanding, we do see that the <b> in "climb" is there for a reason. Both hypotheses draw attention to the etymologically related word "clamber" so we have an easier time remembering this spelling - and we don't pass it off and all those other words we found in the Word Searcher as 'exceptions'.
For many in the SWI FB group where this was discussed, the concept of an orthographic marker may be new, and for some it is a well-established concept. Outside of SWI/real spelling based instruction, however, this concept is essentially absent. Without this concept, we cannot explain the grapheme-phoneme correspondences in words like "two" "people" "please" and countless others.
And this is one of the beautiful things about taking on the practice of  "spelling-out-orthography". It forces us to try to identify the orthographic structures in words. Since countless words have orthographic markers, teachers who don't know about that structure will run into difficulties spelling-out words that have them. That, in turn provoks people like the questioner in this case to ask a question in a community of orthographic scholars. And by asking that question, that community has a reason to bring up this concept for the questioner and others reading the FB string. 
Not to be a car salesment, but I do want to mention that if you've read this long and are interested in how to understand and teach these concepts, I do recommend my 2-session course on "spelling-out-orthography". This process is part of my 5-session course, but we really have time to dive in more deeply with two 2-hour sessions. I do not have a version of this course booked right now. Anyone interested in that course, or another one of my courses can see descriptions at the link below. If you can find one or two colleagues interested in the same course, we can schedule one at a time that fits your time zone. Just email me at  <This email address is being protected from spambots. You need JavaScript enabled to view it.>. 
That course is not just about the practice of  'spelling-out orthography' as a tool for instruction and assessment, I use this practice to teach essential orthographic concepts like orthographic markers, zero allophones, the homophone principle, the way SWI practice is so well alligned with recommendations of cognitive load theory and much more.