repository · survival report #6 · 2026-10-19 · method v4 · unranked
pydantic/pydantic-ai
In Survival Report #6 (2026-10-19, method v4): 584,351 of 689,024 lines introduced by 354 AI-tagged commits are still at HEAD, 84.8 %. These are counts, not grades. No rank, no colour, no verdict, and no comparison with any other repository on this page. Every number links to the audit bytes behind it and the method and its limits are public.
Baseline: this repository's untagged lines
Commits with no machine-readable AI signal, human-written, inline-completed or untagged-agent code alike: 3,960 commits, 2,075,749 lines introduced, 1,667,215 still at HEAD (80.3 % line-weighted, 75.9 % median per commit). Age is the time from a commit to HEAD. The baseline was added in method v3 and is unchanged later; definition at causari.dev/method.
Age-matched: AI-tagged lines 84.8 % against untagged lines in the matched age windows 83.6 %, a gap of +1.2 pts, over 5 age windows holding 100 % of the AI-tagged lines.
| Line age | AI-tagged | commits | Untagged | commits |
|---|---|---|---|---|
| 0–30 d | 86.2 % | 26 | 87.3 % | 930 |
| 30–90 d | 70.8 % | 40 | 85.0 % | 777 |
| 90–180 d | 76.5 % | 102 | 74.4 % | 468 |
| 180–365 d | 87.2 % | 179 | 84.9 % | 509 |
| 365–730 d | 88.9 % | 7 | 60.1 % | 1,248 |
| 730+ d | — | 0 | 34.9 % | 28 |
By agent, in Survival Report #6
Alphabetical. A commit is attributed to the agent its metadata names; one agent per commit. VERIFIED commits only.
| Agent | Commits | Lines introduced | Still at HEAD | Line-weighted | Capped |
|---|---|---|---|---|---|
| claude-code | 338 | 682,483 | 579,683 | 84.9 % | 76.3 % |
| cursor | 3 | 1,156 | 1,013 | 87.6 % | 87.6 % |
| devin | 7 | 3,815 | 2,656 | 69.6 % | 69.6 % |
| github-copilot | 4 | 1,357 | 805 | 59.3 % | 59.3 % |
| openai-codex | 2 | 213 | 194 | 91.1 % | 91.1 % |
History across reports
One row per report this repository was measured in, oldest first; each report measured HEAD as of its own date. The line is the line-weighted ratio on a fixed 0 – 100 % axis; each point is one report.
| Report | Date | AI-tagged commits | Lines introduced | Still at HEAD | Line-weighted |
|---|---|---|---|---|---|
| Survival Report #2 | 2026-09-23 | 264 | 618,017 | 541,400 | 87.6 % |
| Survival Report #3 | 2026-09-28 | 333 | 682,096 | 592,763 | 86.9 % |
| Survival Report #4 | 2026-10-05 | 347 | 684,751 | 581,635 | 84.9 % |
| Survival Report #5 | 2026-10-12 | 353 | 688,758 | 584,568 | 84.9 % |
| Survival Report #6 | 2026-10-19 | 354 | 689,024 | 584,351 | 84.8 % |
Reproduce
One command, any machine, no account: the same bytes the report was built from. The exact bytes behind this page are repos/pydantic__pydantic-ai.json of Survival Report #6. Open an issue with your JSON if it differs.
re audit pydantic/pydantic-ai --json # method v4, full cloneBadge
Text only, one colour, no red and no green. It follows the latest report and is cached for one hour. badge-dark.svg is the same badge as paper on ink.
[](https://causari.dev/r/pydantic/pydantic-ai/)Coverage
- Method v4, causari 0.4.1. Detection from commit metadata only; survival from
git blame -w -M -Cat HEAD. - Full clone: yes. Method v4 refuses shallow clones; this row comes from a complete history.
- Sample: 354 AI-tagged commits; the sample floor is 5. At or above the floor, so the ratio is aggregated into the report.
- PROBABLE: No PROBABLE commits: every counted commit carries machine-readable AI authorship metadata.
- Cap: 3,781 lines per commit in this repository; the capped ratio is 76.2 %.
.git-blame-ignore-revs: not present in this repository.