abcop
A must-have gate for AI-written code. Agents ship faster than humans can re-read; abcop keeps that code understandable by gating function and module ABC complexity so every unit fits a human head and an LLM context window. Diagnostics are also a hook for automated refactoring — extract a method, split a module — so CI rejects bad growth and points agents at concrete maintainability fixes.
One self-contained binary across Ruby, Rust, Python, Go, PHP, Java, C#, Dart, JavaScript, TypeScript, C, C++, Objective-C, Swift, Solidity and Zig — no runtimes, no plugins, no per-language installs. Written in Rust for speed: one parse per file, one walk per metric, grammars compiled in; whole trees in milliseconds.
Install
macOS (.tar.gz from GitHub Releases; binary + man):
# Apple Silicon — use the *-x86_64-apple-darwin.tar.gz asset on Intel Macs
Ubuntu / Debian (.deb from GitHub Releases; amd64, Ubuntu 22.04+ / Debian bookworm+):
# example for v0.17.3 — use the asset name from the release page
Example run
lib/sinatra/base.rb:1254:0: C: Metrics/AbcSize: Assignment Branch Condition size for `error_block!` is too high. [<7, 14, 9> 18.06/17]
src/main.rs: W: Metrics/ModuleAbcSize: Assignment Branch Condition size for module is too high. [<80, 200, 60> 228.04/120] -- extract a coherent subunit
132 files analysed in 0.09s, 7 abc offenses, 0 used-once offenses, 0 never-used warnings, 14 module-abc warnings
Why ABC, not line counts
Line counts lie. The ABC metric
(Jerry Fitzpatrick, C++ Report, June 1997) counts what does work —
assignments (A), branches (B), conditions (C) — as sqrt(A² + B² + C²).
Fitzpatrick defined both method and module scope; abcop gates both:
| Rule | Severity | Meaning |
|---|---|---|
Metrics/AbcSize |
C | function ABC above --max-abc (default 17) |
Metrics/ModuleAbcSize |
W | module ABC above --max-module-abc (default 120) |
UsedOnce |
W | local written once, read once — consider inlining |
NeverUsed |
W | local written, never read |
A sparse wrapper and a dense god-object can share a line count; ABC separates them. Complexity is the gate — never a line budget. UsedOnce / NeverUsed are secondary.
Built for AI workflows
- Understandable by construction. AbcSize 17 and ModuleAbcSize 120 keep every unit inside a human head and an LLM context window. Models edit small units reliably; when something breaks, the blast radius is tiny.
- Refactoring hook. Exit code
1plus stable JSON/JSONL diagnostics (rule,score,vector,message) feed agents and scripts: split oversized modules, extract hot methods. Each finding is an actionable maintainability step, not a style nit. - Fast enough for every push. Sub-second whole-tree scans — including code an agent will never re-read tomorrow.
- Signal only. No formatting or style cops. Vendored, generated and
test trees are skipped by default. Two knobs:
--max-abcand--max-module-abc.
Inspired by RuboCop (Ruby AbcSize
parity and # rubocop:disable directives) and
lizard (one tool, many languages).
Ruby's counting matches RuboCop 1.89 byte-for-byte
(scripts/compare_parity.py).
Advantages
- ~850k–940k LOC/s per core (Apple M1 Pro). RuboCop's lib tree
(943 files, 110k LOC): 0.13 s — ~39× faster than
rubocop --only Metrics/AbcSize(cache off), ~140× a full rubocop run. - One binary, zero deps — grammars and an embedded result cache
(
cache.redb) compiled in; works on clean CI images with no language toolchains. - Deterministic — same findings every run; text / JSON / JSONL stream
as files finish;
--sort-by-scorebuffers for worst-first emit.
Usage
| Option | Default | Meaning |
|---|---|---|
[PATH]... |
auto | targets; omitted → scope selection |
--max-abc N |
17 |
function ABC ceiling |
--max-module-abc N |
120 |
module ABC ceiling |
--only abc|used-once|never-used |
all | single check |
--full |
off | whole production tree (default skips stay on) |
--everything |
off | no gitignore / hidden / vendored pruning |
--format text|json|jsonl |
text |
CI-friendly output |
--sort-by-score |
off | highest ABC first |
--mr |
off | MR scope (uncommitted + branch vs base) |
--uncommitted |
off | working-tree + index + untracked vs HEAD only |
--no-cache |
off | skip on-disk cache |
--dump-tree FILE |
— | debug syntax tree |
Exit codes: 0 clean, 1 findings, 2 usage error.
JSON diagnostics include file, line, column, severity, rule,
message, plus score / vector for ABC rules:
Scope selection
Named paths — those targets only.
Omitted — narrowest useful scope, announced on stderr:
- uncommitted work vs
HEADif the tree is dirty - else current MR (
--mrforces this) - else full tree (outside a repo)
Default walks prune test/fixture trees, vendored/build output
(vendor/, node_modules/, target/, …), db/migrate/, route tables
(config/routes.rb, config/routes/*.rb), and generated names
(*.min.js, *_pb.go, …). Name a path explicitly to scan it anyway.
Third-party, route-table, and fixture paths are never scoped review
surface — a diff through vendor/ or tests/fixtures/ does not make
that material owned code.
Scoped ModuleAbcSize re-sums only methods that intersect the diff and
compares that total to --max-module-abc (default 120; untracked =
every method). A small patch into an oversized legacy file stays quiet
unless the touched methods themselves exceed the ceiling; AbcSize still
reports any changed method over --max-abc. Full scans (--full,
--everything) report every production module over the ceiling;
ModuleAbcSize still exempts test trees on full scans (scoped runs can
flag them when changed methods sum over the limit). UsedOnce /
NeverUsed always follow the changed lines.
On a dirty tree the bare default is uncommitted-only; --mr takes the
full branch union. Commits straight to main use a 36-hour window when no
branch base applies. --uncommitted fails outside a repository instead
of silently widening.
Directives
Everywhere except Rust, RuboCop-style # / // suppressions work
(trailing and block; bare Metrics allowed). rubocop:disable-next is
ignored, matching rubocop.
# rubocop:disable Metrics/AbcSize
...
end
Caching
Content-addressed cache under $XDG_CACHE_HOME/abcop (or ~/.cache/abcop;
override with ABCOP_CACHE_DIR). Warm reruns ~5× faster. Keys cover
contents, version, rule revision, threshold, checks and path — no
cross-project collisions. Auto-pruned to 20 000 entries; --no-cache
disables. Nothing is written inside the project.
Supported languages
Sixteen languages, four rules each, one CI gate.
| Language | Files | Notes |
|---|---|---|
| Ruby | .rb .rake .ru .gemspec, Gemfile, Rakefile, … |
RuboCop-parity AbcSize |
| Rust | .rs |
|
| Python | .py .pyi .pyw |
|
| Go | .go |
|
| PHP | .php |
|
| Java | .java |
|
| C# | .cs |
|
| Solidity | .sol |
|
| Dart | .dart |
|
| Zig | .zig |
|
| JavaScript | .js .mjs .cjs .jsx |
|
| TypeScript | .ts .tsx .mts .cts |
|
| C / C++ | .c .h .cc .cpp .cxx .hpp .hxx .hh |
.h via C++ grammar |
| Objective-C | .m .mm |
|
| Swift | .swift |
Named declarations are measured units; anonymous function-likes roll into
the enclosing unit; nested units never double-count. C-family: file-scope
globals are out of single-file reach; loop-head locals are protocol;
field writes also read the object; export-macro class forms
(class UTIL_EXPORT Foo) are skipped for variable rules.
Benchmarks
Apple M1 Pro, warm cache:
| Corpus | Size | abcop | rubocop --only AbcSize¹ |
Full rubocop |
|---|---|---|---|---|
| rubocop/lib | 943 files, 110k LOC | 0.13 s | 5.1 s (~39×) | 18.3 s |
| cargo registry sample | 6,603 files, 2.6M LOC | ~2.8 s | — | — |
¹ --cache false — rubocop's own cache otherwise makes repeats look free.
Development
Shared engine: src/scope_model/ (backend, walk, eval); language
collectors sit beside each backend (clike/, sollang/, …). Probe
grammar nodes with abcop --dump-tree FILE. Parity:
scripts/compare_parity.py.
License
GNU GPL v3 or later (SPDX: GPL-3.0-or-later). See LICENSE.
Copyright © 2026 Peter Adrianov. All rights reserved.