crc-checksum-integration

# CRC Checksum Integration ## Overview Use this skill to integrate CRCs and checksums by making the exact algorithm parameters explicit. Most CRC bugs come from two sides using the same name for different init, reflection, xorout, byte order, or covered range. ## When To Use Use this skill when: - The user needs CRC-8, CRC-16, CRC-32, additive checksum, Fletcher, Adler, or protocol frame validation. - The issue involves mismatched CRC values, boot image validation, Modbus, CAN, UART packets, storage records, or OTA packages. - The target uses hardware CRC units, DMA, endian-sensitive protocols, or streaming updates. Do not use this skill for cryptographic authentication. Use `mbedtls-integration`, `tinycrypt-integration`, or `micro-ecc-integration` when adversarial tampering matters. ## First Questions Ask for: - Algorithm name plus polynomial, width, init, refin, refout, xorout, and check value if known. - Exact byte range covered, header/trailer inclusion, padding, and endian order. - Whether the implementation is table-driven, bitwise, hardware accelerated, or streaming. - Known-good sample input and expected output from the peer device or specification. - Current mismatch and where it is observed. ## Integration Checklist 1. Write the full CRC tuple. Do not rely on names like CRC-16 unless the parameters are specified. 1. Verify with a golden vector. Test `123456789` or a protocol-provided frame before using live data. 1. Freeze byte coverage. Define...