1. Sraffa (and more here and here) wrote the seminal paper on the size of the firm under perfect competition. If contracts are perfect, what does a firm mean anyway? He doesn’t quite get to that point, but he did understand that if all the cost curves and demand curves are horizontal, we cannot say how small or how large an individual firm will be. For 1926 this article was amazing.
2. He wrote an excellent 1932 review of Hayek’s business cycle theory. First he questioned the idea of a "natural rate of interest" as a base from which the market rate might deviate. He pointed out that there is an entire schedule of natural rates. His deeper point was that monetary policy-induced resource shifts were not necessarily unstable and due for costly reversal. The newly-favored sectors might simply keep the resources they had captured through seigniorage, "forced savings," and other non-neutral monetary effects. Oddly the economic pessimists, usually on the left, have forgotten these points in recent times.
3. He wrote a beautiful introduction to the collected works of David Ricardo.
4. In 1960 he published the neo-Ricardian cult tract The Production of Commodities by Means of Commodities. Sadly this book is almost entirely misleading. It tries to resurrect the Ricardian idea that costs alone determine price (or perhaps I should say it cryptically hints at that idea without actually defending any real propositions at all).
At first glance cost-based theories seem plausible. Draw a horizontal cost curve, as would arise from constant returns to scale, and demand will not matter for price. Are constant cost such an unreasonable assumption? Can we not in principle multiply every economic input and receive proportionate outputs? Not really. There are almost always fixed factors and the "pure multiplication" thought experiment is not very relevant for price in the real world. Both Marshallian scissors matter.
Now this was not Sraffa’s argument, but Sraffa’s argument was far more complicated and also worse. It relied on aggregation theorems which simply didn’t stick, and Mark Blaug tore it apart.
Sraffa also had important relationships with Gramsci and Wittgenstein, but those are topics for other posts, not necessarily by me.
Overall the guy’s output was quite sparse. He was brilliant before the War but his major book was headed in the wrong direction altogether and it should be taught as a mistake, not an alternative.