ActuallyTaylor presents Strata’s published GitHub CSV as a mined list of 1,325 AI-assisted repositories, potentially providing an inspectable corpus for repository-level analysis of AI-assisted development.
state: seedheat: lowuncertainty: mediumnovelscott: lowai-adoption software-provenanceActuallyTaylor
What is this?
The case describes ActuallyTaylor presenting a Strata-published GitHub CSV as a mined collection of 1,325 AI-assisted repositories, with repository metadata and an assistance-style field. None of the supplied web results identifies ActuallyTaylor or Strata or verifies the CSV, its size, schema, or mining methodology. The snippets establish only adjacent activity—GitHub repositories tagged as AI-assisted and a separate framework claiming to detect assistance through commit-message patterns—so the proposed corpus’s availability and suitability for repository-level analysis remain unverified.
Why it matters to Scott
No meaningful intersection established: repository metadata and assistance-style labels do not supply the production-time lineage Scott argues for in Provenance-Coupled Work or the work traces his Second-Order Learning concept mines. The corpus and mining method remain unverified, no supplied hit establishes a use for it in his active projects, and the radar hits do not track this same development.
queries asked of Scott's wikis
- AI adoption measurement repository mining
- software provenance AI-generated code attribution
- coding agent evaluation real-world repository corpora
- AI assistance styles human agent collaboration
- self-reported AI usage detection sampling bias
Measured heat
now 0 pts/hpeak 0 pts/hcomments 0/hpeers p14momentum: steady2 platformsage 649h
points/hour across evidence · reading as of 2026-10-12 02:59:37.977291+11:00 · deterministic, not a model opinion
How the heat travelled
pace: p56 vs 1032 stories at the 336h mark (now 649h old) — ahead of un-ccw-human-review-removal (1.0x), behind 37signals-agent-driven-default (1.0x)
Evidence (2) — ⭐ canonical anchor
| source | object | author | score | comments |
|---|
| 🟧 hn | A list of 1,325 AI assisted repositories, mined from GitHubRetrieved article excerptOpen article · Retrieved 2026-09-14T15:23:23.522639+00:00 [ActuallyTaylor](https://github.com/ActuallyTaylor)
/
**[strata](https://github.com/ActuallyTaylor/strata)**
Public
- [Notifications](https://github.com/login?return_to=%2FActuallyTaylor%2Fstrata) You must be signed in to change notification settings
- [Fork
0](https://github.com/login?return_to=%2FActuallyTaylor%2Fstrata)
- [Star
0](https://github.com/login?return_to=%2FActuallyTaylor%2Fstrata)
## FilesExpand file tree
main
/
# ai-assisted-repositories.csv
Copy path
BlameMore file actions
BlameMore file actions
## Latest commit
## History
[History](https://github.com/ActuallyTaylor/strata/commits/main/paper/data/large/datasets/ai-assisted-repositories.csv)
History
1325 lines (1325 loc) · 277 KB
main
/
# ai-assisted-repositories.csv
Copy path
Top
## File metadata and controls
- Preview
- Code
- Blame
1325 lines (1325 loc) · 277 KB
[Raw](https://github.com/ActuallyTaylor/strata/raw/refs/heads/main/paper/data/large/datasets/ai-assisted-repositories.csv)
Copy raw file
Download raw file
Edit and raw actions
| 1 | repo\_id | language | stars | url | description | default\_branch | assistance\_style |
| --- | --- | --- | --- | --- | --- | --- | --- | | ActuallyTaylor | 11 | 14 |
| 🟧 echo.github ⭐ | Publishes an AI-assisted-repository CSV with repository identifiers, language, stars, URL, description, default branch, and assistance-style | ActuallyTaylor | — | — |
Interpretation history
2026-09-14T15:31:12Z
grounded: novel/low — No meaningful intersection established: repository metadata and assistance-style labels do not supply the production-time lineage Scott argues for in Provenance
2026-09-14T15:28:43Z
case created — The accessible first-party dataset is a concrete artifact, although its classification quality and claimed repository count are not established by the retrieved header and line count.
Decision trace
- 10-10 05:16review_dormantscheduled targets exhausted or 28 quiet days
- 10-10 05:16drop_targetsquiet through full ladder or over cap 8
- 09-15 02:33review_screenThe new comments add discussion, questions, and an unsubstantiated security concern but no verified change to the dataset, repository count, classification quality, or analysis results.
- 09-15 02:20sensor_dirtycomment_update
- 09-15 01:31groundNo meaningful intersection established: repository metadata and assistance-style labels do not supply the production-time lineage Scott argues for in Provenance-Coupled Work or the work traces his Sec
- 09-15 01:28createThe accessible first-party dataset is a concrete artifact, although its classification quality and claimed repository count are not established by the retrieved header and line count.