I've updated docstrings for several Minimal MeTTa functions to offer more insight on their behavior at least as documented in https://github.com/trueagi-io/hyperon-experimental/blob/main/docs/minimal-metta.md and as explained by Vitaly on the Mattermost.

minimal_metta_tests.metta I intend to include basic, if you will minimal test cases that both provide behavioral coverage and augment the documentation. (I have not completed it for all of Minimal MeTTa yet, but the current tests are a good start.) They are based on my understanding of the Minimal MeTTa specification above and the Hyperon implementation's behavior. The Hyperon implementation currently passes all of these tests save for some of the collapse-bind tests, which appears to be an issue with equality checking for assertEqualToResult rather than an issue with the test: the results returned match the test specification exactly. For example, in the Metta playground this fails but should succeed:

(= (bar A) (input A))
(= (bar B) (input B))
(= (foo $x) (bar $x))

!(assertEqualToResult (chain (collapse-bind (foo $x)) $a (quote $a)) ((quote (((foo $x) {  })))))

[(Error (assertEqualToResult (chain (collapse-bind (foo $x)) $a (quote $a)) ((quote (((foo $x) { }))))) 
Expected: [(quote (((foo $x) { })))]
Got: [(quote (((foo $x) {  })))]
Missed result: (quote (((foo $x) { }))))]

Still, all in all the Hyperon implementation is currently doing quite well on test body I've created. Mettalog is not doing so well, only passing 10/36 of these new tests atm.