Rank-metric cryptography can produce much smaller public keys than the comparable Hamming-based analogues, but the history of rank-metric signatures highlights a pitfall: adding too much structure to the code has repeatedly opened the door to devastating algebraic attacks. In this talk, we revisit the conservative assumption of designing signatures directly from the Rank Syndrome Decoding problem, relying on the hardness of decoding random rank codes. We trace two cryptographic lineages stemming from Stern’s zero-knowledge proof protocol — Chen’s rank-metric adaptation and the five-pass Hamming CVE scheme — then combine these frameworks with modern Fiat–Shamir optimizations to obtain RaCSS. At 128-bit security, RaCSS achieves a public key of roughly 40 B and signatures below 10 kB. We leverage this signature scheme to develop the first known linkable ring signature native to the rank metric, an advanced functionality permitting a user to sign anonymously on behalf of a group, with competitive sizes for practical rings. Finally, we show that an earlier published rank-CVE signature proposal, Bellini et al., admits a universal forgery attack wherein an adversary can generate valid-appearing signatures without a private key, emphasizing the import of sound rank-metric signature schemes.