I guess agricultural loans as percentage of agricultural value added and access to finance can serve as a nice proxy for rural financial development of rural economy.
Causality could be bi-directional in this case. Have you tried to look at time series data for your country or the country of interest in your research? Or perhaps panel data or cross-sectional data for a number of countries of interest to you? You can establish causality using the Granger causality test for instance if you have the data. You can look at causality between GDP and M2/GDP or M3/GDP.
On the second question, I agree with Balazs that financial development would be roughly gauged from access to various financial instruments, that is, offering savers and borrowers a broad range or financial instruments, different in terms of liquidity, yields, maturities, etc, etc.
In case you are still interested in this question, you will probably find a bidirectional causality between financial development and growth, as Barnabas ponted out. However, the key is what relation are you interested in, so as, probably, inducing further financial development will translate into economic growth and viceversa.
May be the references in some papers that I coauthored some time ago could be of help for you. As to the assessment of financial development, you can adapt the measure for financial development that we proposed by considering just the financing related to the agricultural sector. In this sense, I would also include micro-credits and informal credit where data were available. Besides, if flow of funds data were available too, the information content on the liability side for the agricultural sector (if available) could be of great help, but in this sense, it is not easy to find such statistics. (Please see: “Why do countries develop more financially? the role of the central bank and banking supervision", with A.García Herrero and S.Gallego (2003) for a comprehensive measure of financial development).
As for the relation relation between financial development and growth, you could have a look to "Does a balanced financial structure matter?” with A. García Herrero (2008), where there are a few references that could help in your research.
At first you must have some level of development to start with then by time a country will grow, this means development will cause growth first. Then after building some momentum economic growth will be the main cause for more development later, only at this time the relation will converge to tow way causation,