vercel-labs/agent-desktop

Code health of vercel-labs/agent-desktop

Repowise scores vercel-labs/agent-desktop at 9.3 out of 10 for code health, which reads as excellent, alongside maintainability 9.4 and static performance 9.9 on the same scale. There are 3,901 open findings across the three pillars, and the lowest-scoring file is shell-probe.cs at 1.0. Scores are computed from complexity, duplication, churn and git fix history, not from a linter.

Findings by severity

113 critical, 216 high, 1,200 medium, 2,372 low.

Hotspots

309 files are flagged as hotspots. The top 5:

crates/core/src/snapshot/mod.rs: 10 commits in the last 90 days from 1 contributor with 5 recent bug fixes
crates/ffi/include/agent_desktop.h: 10 commits in the last 90 days from 2 contributors with 5 recent bug fixes
crates/macos/src/tree/resolve/mod.rs: 8 commits in the last 90 days from 1 contributor with 5 recent bug fixes
crates/macos/src/actions/dispatch.rs: 6 commits in the last 90 days from 1 contributor with 3 recent bug fixes
src/dispatch/mod.rs: 6 commits in the last 90 days from 2 contributors with 2 recent bug fixes

Lowest-scoring files

probes/windows/26-shell-surfaces/shell-probe.cs: 1.0 out of 10
npm/scripts/postinstall.js: 3.0 out of 10 with 4 bug fixes in the last 180 days
tests/e2e-linux/scenarios/ref_contract.py: 3.0 out of 10
scripts/check_architecture.py: 3.3 out of 10 with 1 bug fix in the last 180 days
crates/macos/src/system/cursor_overlay/cursor_overlay_bridge.m: 3.5 out of 10 with 2 bug fixes in the last 180 days

Checked against bug fixes

Of the 20 files Repowise scores lowest, 14 had a bug fix in the last 180 days. Across all 3,125 scored files, 17.2% had a bug fix in that window.

Dead code and refactoring

Repowise found 8 unused exports. 98 refactoring targets across 98 files would raise the score.

More on agent-desktop: summary and FAQ, codebase overview, architecture.