Review economics
kvcache-ai/Mooncake
Read from the 100 most recently merged pull requests ·
Pull requests with attributed agent authorship took only 0.7× the reviews submitted of the rest.
≥10%
Attributed
a floor
13
Attributed PRs
87
Other PRs
100
Read
Attributed — the latest 5
#3296 [TE] EFA: derive WR/CQ pacing depths from the libfabric provider
1 reviews · 24h to merge
#3237 fix: add CUDA stubs dir to link_directories for no-driver builds
6 reviews · 2.5d to merge
#3254 [Store] Fix processing_keys double-erase UAF in MetadataAccessorRW
2 reviews · 40h to merge
#3214 [Store][TE] fabric_mem best-effort alloc via percentile ladder
8 reviews · 4.9d to merge
#2420 [Docs] [4/N] Refactor Readme: simplify hardware and build sections
2 reviews · 2h to merge
The rest — the latest 5
#3282 [TENT] Add native UB foundation and URMA resource management
4 reviews · 46h to merge
#3272 [TE] Support DMA-BUF in IBGDA Device API
30 reviews · 2.1d to merge
#3309 fix: resolve nightly CI failures
6 reviews · 2h to merge
#2955 [TransferEngine] Skip the CUDA pointer probe on GPU-less hosts
4 reviews · 20.8d to merge
#2962 [TransferEngine] Make single unregisterLocalMemory best-effort
2 reviews · 20.7d to merge
How this was measured
10% of merged commits carry agent attribution — a floor, not the share; tools that only complete code inline leave no commit trail, so the unattributed side includes AI-assisted work; 25 PRs came off an agent branch with no commit attribution, so the real share runs higher; reads 100 of 1844 merged PRs — the rest have not been analyzed yet.
Detected: OpenAI Codex, Claude Code, GitHub Copilot coding agent, Cursor agent, Gemini CLI.
“Attributed” means a commit carried an agent’s signature — a co-author trailer, an agent commit identity, or an agent bot account. Tools that only complete code inline leave no such mark, so the other column is “rest”, not “human-written”. Full method and its limits
The badge reads this repo’s current report, so it follows the number.
Tell me when this moves
We re-read kvcache-ai/Mooncake weekly and email only when the number changes materially.
Read your own repos
This one is public. For a private repo, run the same read locally through your own GitHub credentials — nothing is installed and nothing is sent to us.
npx @ambera/review-taxWant this continuously on kvcache-ai/Mooncake — each pull request paired to the task it came from, and the work graded from its review loop rather than its diff size? Claim this repo in Forge
Browse every repo people have read · Read a different repository