dietrichgebert / ponytail
markdownmaina9457787.9K linessynced 1mo ago
Code health
8.7out of 10Excellent
This codebase scores 8.7 out of 10 on defect risk, which we rate excellent. It also scores maintainability 9.3 and static performance risk 9.8 out of 10. The three are scored separately and never blended into one number. The files you change most are the weak spot: 8 of 126 files are git hotspots, and they average 6.2.
- Lines of code
- 7.9K
Recent activity
All commitsCommits
No commits indexed yet.
Decisions
- Concatenate agent source files with headers for judging inputproposed
- Judge only source files; exclude tests from over-engineering scoringproposed
- Use an LLM judge for over-engineering with an auditable rubric and self-validationproposed
- Treat device as non-ideal: preserve calibration knobs/flags and require actionable calibration mentionproposed
- Behavior gate requires produced refined behaviors via one actionable check per probeproposed
- Write only what the task needs (no golfing)proposed
Needs attention
55 open- medium severity. ProposedConcatenate agent source files with headers for judging inputAuto-proposed decision awaiting review
- medium severity. ProposedJudge only source files; exclude tests from over-engineering scoringAuto-proposed decision awaiting review
- medium severity. ProposedUse an LLM judge for over-engineering with an auditable rubric and self-validationAuto-proposed decision awaiting review
- medium severity. ProposedTreat device as non-ideal: preserve calibration knobs/flags and require actionable calibration mentionAuto-proposed decision awaiting review
- medium severity. ProposedBehavior gate requires produced refined behaviors via one actionable check per probeAuto-proposed decision awaiting review
Where the risk concentrates
All 8 hotspotsRanked by prior bug fixes and change frequency, mined from full git history rather than from the code alone.
| File | Churn | Prior fixes | Maintainers | Commits 90d |
|---|---|---|---|---|
| tasks.pybenchmarks/agentic | 97.8th | 1 | 2 | 3 |
| run.pybenchmarks/agentic | 95.6th | 2 | 2 | 3 |
| ponytail-activate.jshooks | 93.3th | 6 | 7 | 11 |
| correctness.jsbenchmarks | 91.1th | 2 | 3 | 3 |
| hooks.test.jstests | 88.9th | 4 | 7 | 9 |
Composition
Open the graphmarkdown 43%javascript 30%json 11%yaml 6%toml 5%Other 5%
Explore this codebase
- Docs0 pages across 58 modules
- ChatAsk this codebase a question and get an answer with its sources
- Files126 files with per-file docs, health and history
- ArchitectureDependency graph, layers, and 28 entry points
- Code healthPer-file scores, 190 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