shinkuan / akagi
Code health
This codebase scores 5.9 out of 10 for code health, which we rate fair. It also scores maintainability 7.7 and static performance 9.7 out of 10. The three are scored separately and never blended into one number. Risk is concentrated, as it usually is: 47 of 352 files are git hotspots, and they average 3.7 — which is where the fixes pay off most.
Do next
4 things worth doing this quarter.
Worth planning
Break up load_config (CCN 13, 60 lines)
src/config/mod.rs:167
load_config nests its control flow deeply; 36 files import it, changed 11 times in 90 days.
- gain
- up to +2.2 health on this file(inferred)
- finding
- Nested complexity
- size
- CCN 13, 60 lines, nests 8 deep
- severity
- critical
Medium effort, worth a checkFlatten the nesting in plan
src/autoplay/majsoul/mod.rs:48
plan nests its control flow deeply, and part of it is duplicated, so splitting it also removes a copy; 7 files import it, changed 20 times in 90 days.
- gain
- up to +1.1 health on this file(inferred)
- finding
- Nested complexity
- size
- CCN 27, 129 lines, nests 5 deep
- severity
- high
Medium effort, worth a checkDelete 35 unused symbols and files (705 lines)
Nothing in the graph reaches them, across 18 files; every reader and agent pays to skip them.
- symbols
- 35
- lines
- 705
- safe to delete
- Yes(inferred)
Medium effortBreak up aggregate (CCN 50, 264 lines)
src/history/aggregator.rs:102
aggregate nests its control flow deeply; 5 files import it, changed 4 times in 90 days.
- gain
- up to +1.5 health on this file(inferred)
- finding
- Nested complexity
- size
- CCN 50, 264 lines, nests 6 deep
- severity
- high
Medium effort, worth a check
Improve what Repowise can see
Add a test coverage report
Without one, Repowise cannot tell tested code from untested code, so every test-related action says “unknown”.
- line coverage
- Unknown, no report
Small effort
- Lines of code
- 76.2K
Recent activity
All commitsCommits
- 588c0dfcstyle: apply rustfmt to #257 changes1mo ago
- 77cd181drefactor(autoplay): remove dead synthetic-reach Path B; decline tile-less reach1mo ago
- 13af54fffix(bot): resolve plain-mjai riichi discard via autoplay reach follow-up1mo ago
- 535a4dbadocs: chankan window, discard guard, and blind-riichi refusal1mo ago
- 4a3afc5ffix(capture): subscribe the pause listener before arming the interceptor1mo ago
- ca67bca3fix(tenhou): resolve the thrown tile before acting, and to the exact copy1mo ago
Decisions
- fix(capture): unmute CDP Chromium by defaultproposed · 1mo ago
- fix(tenhou): use 4-cycle seat math for sanma gamesproposed · 1mo ago
- Read active bot from live config at each game startproposed · 1mo ago
- Track autoplay CDP page handle by browser tab, not WebSocketproposed · 1mo ago
- Emit start_kyoku.scores in absolute-seat order for Tenhouproposed · 1mo ago
- Answer forced moves locally to avoid unnecessary API callsproposed · 1mo ago
Everything we found
1,447 open
Every detector's open count by area, each with its worst example. An inventory, not a to-do list: the actions above are the part worth doing.
- worst item is critical severity. Code health924Untested hotspot · src/lib.rs12 findings · 9.6 deducted
- worst item is medium severity. Refactoring130main+1.5 health, S effort · native_bot/src/bin/extract.rs
- worst item is medium severity. Decisions67src/autoplay/majsoul/mod.rs20 awaiting review · 47 hotspots ungoverned
- worst item is medium severity. Ownership291frontend/src/types.ts100% single-owner concentration, 22 commits in 90d
- worst item is low severity. Dead code35CardFooterUnreachable function (11 lines)
Composition
Open the graphExplore this codebase
- Docs507 pages across 8 modules
- ChatAsk this codebase a question and get an answer with its sources.
- Files352 files with per-file docs, health and history
- ArchitectureDependency graph, layers, and 7 entry points
- Code healthPer-file scores, 1,017 open findings, coverage and refactoring targets
- Knowledge graphEntities, communities, and the paths between them
- Change couplingFiles that keep changing together, mined from commit history
- CommitsChange-risk ranked history with agent provenance
- ContributorsBus factor, per-file maintainers, and the human/agent split
- StatsSize class, origin, lifetime churn, rhythm and records
- CostsWhat indexing this snapshot cost, by model and by run.