- Dataset
- 2026-09-18-v3.0
- Ruleset
- b2e8889c66f7
- Scanned
- 2026-09-16
- Window
- February–September 15, 2026
01Result
JuliusBrussee/caveman ↗JuliusBrussee/caveman AI coding statistics: 56%
A small group is using agents. Adoption has spread beyond one person but is not project-wide.
02Drivers
What shapes this result
Mode, timing, and detected signatures show whether the result reflects broad adoption or one concentrated workflow.
Assisted or autonomous
Assisted — a developer’s identity authored the commit and the agent added itself as co-author. Autonomous — a bot identity authored the commit without a human identity attached. Across the whole index, 96% of agent commits are assisted.
Agent share by month
| Month | Agent share | Considered commits |
|---|---|---|
| 2026-04 | 24% | 104 |
| 2026-05 | 78.8% | 33 |
| 2026-06 | 78.3% | 23 |
| 2026-07 | 55.9% | 34 |
| 2026-08 | 44.2% | 217 |
| 2026-09 | 86.2% | 152 |
Peak month: 86.2%. Bars are scaled to that peak, not to 100%.
Which agents
| Agent | Vendor | Commits | Share |
|---|---|---|---|
| Claude Code | Anthropic | 314 | 55.8% |
| Cursor Agent | Anysphere | 2 | 0.4% |
03Evidence
Inspect the matches
Public commit samples let you inspect the signatures behind the aggregate result.
Matched commit samples
ed37ab13on GitHubdaily triage 2026-09-15: shrink compound-word corruption, win32 junction fClaude Code4df4b03bon GitHubchore(release): cut @caveman-ai/cli 1.3.4 on the bin-v1.1.7 runtimeClaude Code8b0c1d36on GitHubchore(release): v2.7.0Claude Code1b7ff8faon GitHubfix(proxy): scope repo intelligence to project source and require direct eCursor Agent902eba34on GitHubfix: opencode .jsonc config, plugin agent discovery, shrink nested descripCursor Agent
04Method
What entered the denominator
The result begins with commit history, removes merges and routine automation, and reports literal agent signatures as a lower bound.
Exclusion ledger
| Commits in window | 672 |
| Merge commits removed | −84 |
| CI / dependency bots removed | −25 |
| Considered | 563 |
Bots are excluded from the denominator entirely. Agent attribution requires a literal supported signature; no code-style inference is used.
05Provenance
Pinned to a source state
This observation names the repository state, release, and detector that produced it. Use these identifiers when reproducing or citing the result.
- Dataset release
- 2026-09-18-v3.0
- Scan date
- 2026-09-16
- Ruleset
- sha256:b2e8889c66f72c18a61839f0bf1a9f77b5481ba2def044dba33c797b2f2bdcae
- Canonical repository
- JuliusBrussee/caveman
- Default branch
- main
- Resolved HEAD
- ed37ab132393899c129bbeef2b9743ff3af19c68 ↗
- Resolved
- 2026-09-16
- Measurement window
- February–September 15, 2026
06Peer context
Position among comparable repositories
The cohort uses the same release and measurement window, then narrows to the repository’s ecosystem and removes low-activity observations.
Midrank percentile; equal shares split the tied positions.
- Repository
- 56%
- Cohort median
- 2.8%
- Share rank
- 7 / 132
Percentile = (repositories below + half of 1 tied repository) ÷ 132.
Continue exploring
Related coding-agent evidence
Run the same detector locally.
Deterministic signature matching; no code-style inference. Squash merges can remove trailers, so this result remains a measured floor.
npx qmmit-cli JuliusBrussee/caveman