A collection of protobuf utilities for working with binary descriptors and wire-format messages.
→ Tutorial: from first decode to schema inference
Lossless, bidirectional converter between binary protobuf wire format and human-readable text. Decodes with or without a schema; encodes annotated prototext back to byte-exact binary. Supports schema inference: given a descriptor DB it ranks all candidate types and picks the best match.
→ See prototext/README.md
Reconstructs compilable .proto source files from binary or text-format
FileDescriptorSet / FileDescriptorProto blobs. Supports proto2, proto3,
and editions. Output recompiles to a descriptor equivalent to the input.
→ See reproto/README.md
Generates pseudo-random binary protobuf instances from a .desc
FileDescriptorSet. Used to populate schema-DB test corpora and as a
quick sanity-check that a descriptor is well-formed.
reproto-instantiate-schema --descriptor-set my.desc --seed 42 -O out/ \
google.type.PostalAddress google.protobuf.Timestamp
Scans binary files (shared libraries, executables, firmware images) for
embedded FileDescriptorProto blobs and extracts them as individual .pb
files, one per proto file descriptor found. The extracted blobs can be
fed directly to reproto to recover .proto source.
protoscan --proto_out extracted/ some_binary
reproto --use-variant descriptor -O src/ -I extracted/ extracted/*.pb
cargo install prototextpip install prototext-reproto # reproto, reproto-instantiate-schema
pip install protoscan # protoscan
pip install prototext-codec # Python extension (lossless codec)
pip install prototext-graph # Python extension (scoring graph)
pip install fdp-scan # Python extension (descriptor scanner)git clone https://github.com/ThalesGroup/prototools
cd prototools
nix-shell # user shell — all tools on PATHMIT — see LICENSES/MIT.txt.