Audited the published part project XC7A50T-1FTG256C (Part UID 705ae7d9-b982-4d06-8722-6dff2d32acfd) without changing its symbol or footprint. The existing project property named exactly Datasheet URL was already populated, so no property mutation was required.

The complete 256-terminal symbol table was enumerated and keyed by physical ball identifier (Pin Number). It was compared ball-for-ball against all 256 entries in AMD's xc7a50tftg256pkg.txt. Duplicate ball IDs, missing official balls, extra symbol balls, and ball/name mismatches were checked.

The symbol uses the correct JEDEC BGA row sequence: A, B, C, D, E, F, G, H, J, K, L, M, N, P, R, T (I, O, Q, and S omitted), columns 1–16.

All terminal designators/names match the official device-specific pinout file. However, terminal descriptions are absent across the symbol; the official pin function is encoded only in the terminal name. Pin grouping/orientation/order metadata is also absent, though explicit custom-symbol coordinates are present.

The project assigns every user/configuration/analog signal terminal Pin Type = Bi-Directional. That is reasonable for general HR I/O but over-broad for dedicated inputs, outputs, and analog pins. Power and ground terminals are consistently Power In.

General HR I/O terminals, including dual-purpose configuration functions embedded in I/O names, remain appropriately represented as bidirectional.

Asset: BGA256C100P16X16_1700X1700X155.kicad_mod

PK053 identifies FTG256 as 17.0 × 17.0 mm nominal, 16×16 balls, 1.0 mm pitch, nominal package ball diameter approximately 0.60 mm, and maximum package height approximately 1.40 mm. The footprint's 0.50 mm feature is the PCB land diameter, not the package ball diameter; it therefore must not be compared as if it were the physical solder-ball diameter. The 0.50 mm land is plausible for a 1.0 mm-pitch BGA, but the project does not record a cited AMD PCB-land-pattern recommendation or assembly-process justification.

The footprint filename suffix X155 suggests a 1.55 mm model/package-height convention, while PK053 reports approximately 1.40 mm maximum package height. The actual STEP model height was not dimensionally extracted by the available structured layout reader, so this is a model-height naming concern, not a confirmed geometry failure.

Every one of the 256 footprint pad labels maps to exactly one symbol Pin Number, and the label set equals the official 256-ball set. No missing, extra, duplicate, or wrongly mapped footprint pads were found.

The part is ball-map complete and internally consistent: all 256 official ball identifiers and names match, all 256 pads exist, and symbol-to-footprint mapping is one-to-one. Package planform and pitch match FTG256. The principal quality gaps are semantic/documentation rather than mapping errors: dedicated JTAG/configuration/analog terminals are typed too generically as bidirectional; all pin descriptions are absent; package metadata says only 256-BGA; and the footprint/model naming hints at a possible height discrepancy requiring direct STEP measurement. No symbol or footprint changes were made, per audit-only scope.