docs/maintainers/excision-v0.8.0-line-count-report.md
Comparison of integration/v0.8.0 (after the excision pass) vs upstream/master at a461b016d.
Lines counted via /tmp/lcount.py against:
crates/, src/, tools/, xtask/, tests/, benches/, fuzz/web/src/docs/Categories:
tests/ directories AND #[cfg(test)] mod blocks within source files/// and //!, TypeScript /** ... */// (non-doc), TypeScript // and /* ... */docs/Excluded: target/, node_modules/, dist/, .git/, tmp/, docs/book/book/ (generated mdbook output), web/src/lib/api-generated.ts (generated TS bindings), docs/book/po/ and po-extract/ (translation artifacts).
| category | master | branch | Δ |
|---|---|---|---|
| logic | 187,521 | 193,324 | +5,803 |
| tests | 118,974 | 123,007 | +4,033 |
| docstrings | 14,765 | 17,100 | +2,335 |
| comments | 7,572 | 8,894 | +1,322 |
| docs | 11,512 | 14,131 | +2,619 |
| blank | 42,590 | 44,138 | +1,548 |
| TOTAL | 382,934 | 400,594 | +17,660 |
The v0.8.0 branch grows the repo by +17,660 lines total, of which:
| component | Δ | share of growth |
|---|---|---|
| logic | +5,803 | 33% |
| tests | +4,033 | 23% |
| docs | +2,619 | 15% |
| docstrings | +2,335 | 13% |
| blank | +1,548 | 9% |
| comments | +1,322 | 7% |
~67% of the growth is non-logic (tests, documentation, comments, blank). The hypothesis that "a lot of the code bloat size is docs/comments/doc strings" is supported: actual production code accounts for one-third of the branch's expansion. The branch ships V0.8.0 schema overhaul + multi-agent runtime (#6272) + typed-family providers + peer-auth refactor, and most of the line count went into documenting and testing those changes rather than implementing them.
| area | master logic | branch logic | Δ |
|---|---|---|---|
| zeroclaw-config | 12,980 | 17,377 | +4,397 |
| zeroclaw-runtime | 38,480 | 39,373 | +893 |
| web | 20,742 | 21,518 | +776 |
| zeroclaw-macros | 912 | 1,581 | +669 |
| zeroclaw-memory | 5,789 | 6,856 | +1,067 |
| zeroclaw-gateway | 8,759 | 9,323 | +564 |
| zeroclaw-channels | 34,556 | 34,576 | +20 |
| zeroclaw-api | 1,511 | 1,536 | +25 |
| zerocode | 704 | 705 | +1 |
| zeroclaw-tool-call-parser | 1,168 | 1,168 | 0 |
| zeroclaw-hardware | 5,760 | 5,760 | 0 |
| zeroclaw-plugins | 954 | 954 | 0 |
| robot-kit | 1,951 | 1,940 | −11 |
| zeroclaw-infra | 1,024 | 1,021 | −3 |
| src (root binary) | 6,326 | 5,956 | −370 |
| zeroclaw-tools | 24,106 | 23,311 | −795 |
| zeroclaw-providers | 17,582 | 16,152 | −1,430 |
| area | master tests | branch tests | Δ |
|---|---|---|---|
| zeroclaw-channels | 24,825 | 27,392 | +2,567 |
| zeroclaw-config | 9,673 | 11,562 | +1,889 |
| zeroclaw-runtime | 30,665 | 31,493 | +828 |
| zeroclaw-memory | 4,217 | 4,801 | +584 |
| tests (workspace) | 8,275 | 8,162 | −113 |
| zeroclaw-providers | 12,140 | 10,838 | −1,302 |
| zeroclaw-tools | 15,831 | 15,446 | −385 |
| src (root binary) | 3,975 | 3,614 | −361 |
| (others < 200 Δ) |
The branch's pre-excision tip would have been roughly +19,200 total (extrapolating from the diff between the excision-pass commits — 200b8d6ec..6473818c2 — which net ≈ −1,540 lines). The excision pass accounts for roughly 8% of the branch's net growth being shaved off.
By category, the excision pass:
docs/maintainers/excision-v0.8.0-incidents.md.The deletions explain the negative deltas in zeroclaw-providers (−1,430), zeroclaw-tools (−795), and the test reductions in providers (−1,302). The +669 in zeroclaw-macros is non-excision V3 work (Configurable derive extensions for the schema overhaul).
| Total branch growth | +17,660 |
| Production logic growth | +5,803 (33%) |
| Excision pass net | −1,540 |
The branch is large because V0.8.0 is a large feature delivery, but the production-code surface grew by roughly one-third of the headline number. The remaining two-thirds is documentation, tests, docstrings, comments, and whitespace — the supporting infrastructure for the feature work, not the work itself.