Every commit scored for change-risk against this repo's own history, so 'elevated' means elevated here rather than on some global curve.
Needs review
44 commits sit in this repo's top risk tercile, which is 34% of the 130scored. The cut is drawn against this codebase's own history rather than a global curve, so a quiet repo still fills its top band, and here it starts at 6.5 out of 10. What pushes a commit up is size and spread together: a large change confined to one area scores below a smaller one scattered across a dozen files.
Commit categories over time, read off the subject line. Fixes carry the accent because that is the series this chart exists to show.
Began docs-led, now leaning fix.
Ranked by change-risk, highest first. Priority is a tercile of this repo's own distribution, so a quiet repo still fills its top band.
| # | Commit | Author | When | Lines | Risk | Top driver |
|---|---|---|---|---|---|---|
| 1 | b8d6aa7efeat(benchmarks): agentic LOC + safety benchmark answering #126 (#158) | DietrichGebertclaudeassisted | 1mo ago | +1.6K -6 | 100%Elevated | more lines added than baseline |
| 2 | 45f7d2f8Fix/examples issue 127 (#131) | DietrichGebertclaudeassisted | 1mo ago | +1.2K -261 | 99%Elevated | more lines added than baseline |
| 3 | caf138dfbenchmarks: fix correctness gate + robustness audit (#65) (#83) | DietrichGebertclaudeassisted | 1mo ago | +578 -4 | 98%Elevated | more lines added than baseline |
| 4 | b345e493examples: add 6 new over-engineering survivors + platform-native guide (#109) | Jesus cornelio | 1mo ago | +437 -0 | 96%Elevated | more lines added than baseline |
| 5 | 6d990f8cfeat(benchmarks): add correctness assertion (#31) | at384 | 1mo ago | +472 -0 | 96%Elevated | more lines added than baseline |
| 6 | 8c279cbffeat: add pi extension (#1) | Hannes | 1mo ago | +477 -95 | 96%Elevated | more lines added than baseline |
| 7 | ef604945feat: add ponytail skill, examples, README | Emeriko | 1mo ago | +455 -1 | 96%Elevated | more lines added than baseline |
| 8 | 7a3475c0feat: full plugin integration + cross-agent rules | Emeriko | 1mo ago | +673 -14 | 94%Elevated | more lines added than baseline |
| 9 | dedc97cafix: comprehension-first guard + reuse rung (#245, #217) (#253) | DietrichGebertclaudeassisted | 1mo ago | +700 -83 | 93%Elevated | more lines added than baseline |
| 10 | 41d6c2f7feat: ship ponytail to OpenClaw (ClawHub skill package) (#102) | DietrichGebertclaudeassisted | 1mo ago | +382 -0 | 93%Elevated | more lines added than baseline |
| 11 | 9d0118dfdocs: add Korean README translation (#283) | Usul-Devclaudeassisted | 1mo ago | +278 -1 | 91%Elevated | more lines added than baseline |
| 12 | 7f4dc907docs: add Spanish (LATAM) translation of README (#110) | Jesus cornelio | 1mo ago | +216 -0 | 91%Elevated | more lines added than baseline |
| 13 | 386f9573benchmarks: add local model support and Node version note (#63) | Mandavilli Vijay | 1mo ago | +241 -0 | 91%Elevated | more lines added than baseline |
| 14 | 0ac987f9feat(mcp): add ponytail-mcp, an MCP server for the ruleset (#91) | Uchenna | 1mo ago | +155 -0 | 89%Elevated | more lines added than baseline |
| 15 | 955fff53feat(benchmarks): add completeness judge so LOC wins can't hide under-delivery (#171) | Ben Younesclaudeassisted | 1mo ago | +182 -5 | 89%Elevated | more lines added than baseline |
| 16 | c16f967dfeat: codex support | Paul | 1mo ago | +245 -20 | 88%Elevated | more lines added than baseline |
| 17 | c1c80f3cAdding support for Copilot Marketplace plugin (#47) | Max Felker IIclaudeassisted | 1mo ago | +193 -19 | 87%Elevated | more lines added than baseline |
| 18 | 321a59c8feat: reproducible promptfoo benchmark + 3-model results | Emerikoclaudeassisted | 1mo ago | +221 -74 | 87%Elevated | more lines added than baseline |
| 19 | 1b491415docs: correct cost claim to 42-75% from 30-rep re-verification (#129) | DietrichGebertclaudeassisted | 1mo ago | +171 -11 | 86%Elevated | more lines added than baseline |
| 20 | ae24cd00fix: add uninstall cleanup script for state outside plugin files (#226) (#228) | Isha katiyar | 1mo ago | +123 -0 | 85%Elevated | more lines added than baseline |
| 21 | 983255e2docs: A-F benchmark — v4 beats caveman on every axis | dgebert | 1mo ago | +103 -0 | 85%Elevated | more lines added than baseline |
| 22 | b9fa5644feat: inject ponytail ruleset into subagents via SubagentStart hook (#254) | Shane McCarronclaudeassisted | 1mo ago | +95 -2 | 83%Elevated | more lines added than baseline |
| 23 | 706bd279feat: add a dark-background logo (#43) | DietrichGebertclaudeassisted | 1mo ago | +115 -0 | 82%Elevated | more lines added than baseline |
| 24 | 3b4626a9feat: skill v2 — output cap, reflex ladder, benchmarks | Emeriko | 1mo ago | +127 -49 | 82%Elevated | more lines added than baseline |
| 25 | 17a46601feat: publish ponytail as an installable npm package for OpenCode and Pi (#197) | Dang Trung An | 1mo ago | +107 -9 | 80%Elevated | more lines added than baseline |
| 26 | f3da910bfeat: refine ruleset from a full-project field review (#39) | DietrichGebertclaudeassisted | 1mo ago | +202 -15 | 80%Elevated | more lines added than baseline |
| 27 | 46c5c28bfeat: add OpenCode adapter | DietrichGebert | 1mo ago | +156 -1 | 80%Elevated | more lines added than baseline |
| 28 | b609e019docs: benchmark chart — ponytail vs caveman vs control | dgebert | 1mo ago | +78 -1 | 79%Elevated | more lines added than baseline |
| 29 | fe963caeAdd Modern Web Guidance as the rung-3 lookup for web tasks (#82) | Liad Yosefclaudeassisted | 1mo ago | +76 -0 | 77%Elevated | more lines added than baseline |
| 30 | e782790bfeat(benchmarks): add critic-email task — reproduces the critique's own example (#173) | Ben Younesclaudeassisted | 1mo ago | +57 -3 | 77%Elevated | more lines added than baseline |
| 31 | e01aa900feat: add Gemini CLI support (#25) | Ben Younesclaudeassisted | 1mo ago | +83 -1 | 77%Elevated | more lines added than baseline |
| 32 | 88be9caefeat: add publish-openclaw-skills.js to push skills to ClawHub (#273) | DietrichGebertclaudeassisted | 1mo ago | +76 -1 | 75%Elevated | more lines added than baseline |
| 33 | 15749f7ffeat(skills): add /ponytail-gain measured-impact scoreboard (#108) | Ben Younesclaudeassisted | 1mo ago | +119 -3 | 75%Elevated | more lines added than baseline |
| 34 | 8d5037d9docs(readme): add the agentic benchmark chart (#160) | DietrichGebertclaudeassisted | 1mo ago | +66 -0 | 74%Elevated | more lines added than baseline |
| 35 | 147bcfd6fix: use PowerShell $env: syntax for Windows hook paths (#26) | Ben Younesclaudeassisted | 1mo ago | +50 -2 | 73%Elevated | more lines added than baseline |
| 36 | e89a4e98Add rule copy drift check (#3) | Chindanai Jaiman | 1mo ago | +48 -0 | 73%Elevated | more lines added than baseline |
| 37 | 44babb22fix(benchmarks): portable plugin-dir resolution for agentic arms (#170) | Ben Younesclaudeassisted | 1mo ago | +40 -8 | 72%Elevated | more lines added than baseline |
| 38 | 243a28f1feat: skill v3 — compress SKILL.md 115 to 95 lines | Emeriko | 1mo ago | +52 -54 | 72%Elevated | more lines added than baseline |
| 39 | f02f9424fix: use python3 for correctness checks and add CI (#50) | Christopher Mayfieldcursorassisted | 1mo ago | +51 -2 | 70%Elevated | more lines added than baseline |
| 40 | 215777d8fix: don't embed shell-unsafe install paths in statusline setup nudge (#224) | DietrichGebert | 1mo ago | +43 -12 | 68%Elevated | more lines added than baseline |
| 41 | 53fd1e85fix: only deactivate on a standalone "stop ponytail" / "normal mode" (#162) | Lakshya77089 | 1mo ago | +43 -3 | 68%Elevated | more lines added than baseline |
| 42 | 2e6a9376fix(benchmarks): count unfenced code, ASCII-safe output, refresh llama3.2 results (#67) | DietrichGebertclaudeassisted | 1mo ago | +50 -36 | 68%Elevated | more lines added than baseline |
| 43 | 92efc4a6feat: add ponytail-audit skill (#20) | Alexander Bclaudeassisted | 1mo ago | +61 -5 | 68%Elevated | more lines added than baseline |
| 44 | 9c998437docs: same-model control arm, refresh numbers and chart | dgebert | 1mo ago | +42 -22 | 68%Elevated | more lines added than baseline |
| 45 | 763e04defix: align all version manifests to 4.8.1 + guard against drift (#260, #262) (#270) | DietrichGebertclaudeassisted | 1mo ago | +86 -6 | 65%Typical | more lines added than baseline |
| 46 | a4e5e479docs: re-sync README.es.md to current English (agentic numbers, CodeWhale) (#174) | DietrichGebertclaudeassisted | 1mo ago | +43 -6 | 65%Typical | more lines added than baseline |
| 47 | 084f10fbfix: ship the missing /ponytail-help command on Claude Code and OpenCode (#62) | YIZIHN | 1mo ago | +46 -0 | 64%Typical | more lines added than baseline |
| 48 | d9e1480cfeat: add ponytail-debt skill (#40) | DietrichGebertclaudeassisted | 1mo ago | +63 -4 | 64%Typical | more lines added than baseline |
| 49 | 947f2ff4feat(pi-extension): add status bar indicator for ponytail mode (closes #84) (#275) | Tanmay Garg | 1mo ago | +32 -0 | 62%Typical | more lines added than baseline |
| 50 | c15db8d3fix: stop mode filter stripping rule bullets with a colon | Paul Ogier | 1mo ago | +31 -4 | 62%Typical | more lines added than baseline |
Two views of the same model: where the cuts fall, and what commit shape lands you above them.
Every scored commit, binned on the raw 0 to 10 score rather than the percentile. Percentile ranks are uniform by construction, so that axis has no shape to draw. The dashed lines are the tercile cuts behind each row's priority pill.
The 130 most recent commits, on their own recency sample rather than the feed above: that defaults to risk-sorted, so reusing it would plot only the top tercile and call it the spread. Big and scattered is what the model penalises. Click a dot to open it.
Repowise tracks change history across 45 files in dietrichgebert/ponytail. In the last 90 days 45 files were touched, 118 times in total, most often ponytail-activate.js. Every commit is scored for change risk from its size, spread and the history of the files it touches.
