pinocchio

Add pinocchio, pinocchio-system, pinocchio-token, and bytemuck to Cargo.toml, then define an entrypoint with entrypoint!(process_instruction). Route instructions by a single-byte discriminator, validate accounts manually, and use bytemuck for zero-copy account data access.

Pinocchio reduces compute units by 88-95% and binary size by about 40% compared to Anchor, suiting high-throughput programs like DEXs and orderbooks. Anchor is better for rapid prototyping, complex account relationships, and teams unfamiliar with low-level Rust.

Implement manual checks using is_signer(), is_writable(), owner comparison, and PDA derivation with Pubkey::find_program_address. Common patterns include struct-based context parsing, TryFrom implementations, builder-pattern validators, and require! macros.

Yes, pinocchio-token provides Token-2022 instructions like InitializeMetadataPointer and InitializeTokenMetadata, with support under active development. For third-party programs, build instructions manually with AccountMeta and Instruction, then call invoke or invoke_signed.

Pinocchio does not auto-generate IDLs, so annotate instructions and accounts with Shank's ShankInstruction and ShankAccount derives, then run shank idl to produce idl.json. Use Codama with the generated IDL for client generation.

Bytemuck requires 8-byte alignment, so structs with fields smaller than 8 bytes need explicit padding arrays to reach alignment boundaries. Also verify the discriminator byte and account data length before casting with bytemuck::from_bytes.