cuda-python

Public scan — anyone with this URL can view this analysis. Sign up to track your own repos privately, run scheduled re-scans, and get AI fix prompts via your dashboard.

https://github.com/NVIDIA/cuda-python.git · scanned 2026-05-17 02:50 UTC (14 hours, 33 minutes ago) · 10 languages

289 findings (6 legacy + 283 scanner) 87th percentile · Python · medium (20-100K LoC) Scanner says 93 (lower by 12)

File as issue Image

UNIFIED Repobility · multi-layer engine · AI coders

Complete repo analysis

Last scanned 14 hours, 33 minutes ago · v2 · 147 findings from 2 sources. Findings combine the legacy security pipeline AND the multi-layer engine (atlas, wiring, flows, ranked) AND verified AI agent contributions.

JSON

Severity distribution — click a segment to filter

Active filters: excluding tests × Reset all

Bug-class explainers. Each card groups findings of the same shape — these are the patterns most likely to ship to prod and reappear in future scans unless you systematically fix the cause, not just the instance.

Fragile runtime 8 findings

What it is: Code that runs but breaks under predictable input — division by zero, missing keys, unbounded loops, off-by-one slicing.

Why it matters: Reaches production undetected because happy-path tests pass. First user with a weird input crashes the request.

How AI causes it: AI loves writing the happy path; doesn't probe edge cases unless explicitly asked.

Fix approach: Add property-based tests. Wrap external inputs with explicit validators. Use the framework's typed deserializer (Pydantic, attrs).

8 matching findings on this repo

low Duplicated implementation block across source files cuda_core/build_hooks.py:17
low Duplicated implementation block across source files toolshed/_catalog_writer.py:82
low Duplicated implementation block across source files cuda_pathfinder/cuda/pathfinder/_static_libs/find…:52
low Duplicated implementation block across source files cuda_core/cuda/core/_utils/runtime_cuda_error_exp…:249
medium Network/subprocess call without timeout or try/except — benchmarks/cuda_binding…
medium Network/subprocess call without timeout or try/except — ci/tools/merge_cuda_cor…
medium Network/subprocess call without timeout or try/except — ci/tools/fetch_ctk_redi…
medium Network/subprocess call without timeout or try/except — toolshed/check_spdx.py:…

View all fragile runtime findings →

Legacy markers 22 findings

What it is: TODO, FIXME, XXX, HACK comments. Often indicate a known-broken path the author meant to fix.

Why it matters: Each marker is an unfinished thought. Production code shouldn't ship with debt that's documented but not tracked.

How AI causes it: AI mirrors the style of the codebase, so existing TODOs propagate into new code.

Fix approach: Convert each into a ticket. Delete the comment when the ticket lands. Use a pre-commit hook to block new TODOs without an issue link.

12 matching findings on this repo

low Legacy-named symbol `attr_value_list_v2` in cuda_bindings/tests/test_cuda.py:359
low Legacy-named symbol `stream_legacy` in cuda_bindings/tests/test_cudart.py:254
low Legacy-named symbol `init_v2` in cuda_bindings/tests/nvml/conftest.py:17
low Legacy-named symbol `init_v2` in cuda_bindings/tests/nvml/test_device.py:17
low Legacy-named symbol `device_get_count_v2` in cuda_bindings/tests/nvml/test_init…
low Legacy-named symbol `device_get_count_v2` in cuda_bindings/tests/nvml/test_cuda…
low Legacy-named symbol `device_get_attributes_v2` in cuda_bindings/tests/nvml/test…
low Legacy-named symbol `device_get_count_v2` in cuda_bindings/tests/nvml/test_gpu.…
low Legacy-named symbol `init_v2` in cuda_bindings/cuda/bindings/_test_helpers/arch…
low Legacy-named symbol `init_v2` in cuda_bindings/cuda/bindings/_test_helpers/memp…
low Legacy-named symbol `device_get_count_v2` in cuda_bindings/examples/4_CUDA_Libr…
low Legacy-named symbol `simple_zero_copy` in cuda_bindings/examples/0_Introduction…

View all legacy markers findings →

Commented-out code 26 findings

What it is: Lines of source that were intentionally disabled but never deleted.

Why it matters: Git already remembers history — commented code rots, becomes wrong, and adds noise to diffs.

How AI causes it: AI sometimes comments out broken code instead of fixing it. Reviewers approve out of inertia.

Fix approach: Delete. Trust `git log`. If you really need to remember, save it in a notes file under `docs/`.

12 matching findings on this repo

info Commented-code block (5 lines) in cuda_bindings/build_hooks.py:4
info Commented-code block (5 lines) in cuda_bindings/tests/test_cuda.py:241
info Commented-code block (5 lines) in cuda_bindings/tests/nvml/test_init.py:22
info Commented-code block (8 lines) in cuda_bindings/tests/nvml/test_pynvml.py:178
info Commented-code block (6 lines) in cuda_bindings/examples/0_Introduction/vector_…
info Commented-code block (7 lines) in benchmarks/cuda_core/benchmarks/bench_launch.…
info Commented-code block (7 lines) in cuda_pathfinder/tests/test_search_steps.py:151
info Commented-code block (6 lines) in cuda_pathfinder/cuda/pathfinder/_dynamic_libs…
info Commented-code block (6 lines) in cuda_pathfinder/cuda/pathfinder/_dynamic_libs…
info Commented-code block (6 lines) in cuda_pathfinder/cuda/pathfinder/_utils/find_s…
info Commented-code block (6 lines) in cuda_core/tests/test_memory.py:1045
info Commented-code block (8 lines) in cuda_core/tests/test_launcher.py:266

View all commented-out code findings →

{# ── 2026-05-17 Round 14: AI-agent bridge footer ────────────────────── Discoverability: the /agents/voting/ guide + MCP manifest exist but aren't linked from anywhere users actually land. Small, opt-in footer. #}

For AI agents: Voting guide (TP/FP) MCP manifest Stdio wrapper SARIF Integrate Findings queue Vote TP/FP on findings to calibrate the engine.

For AI agents + API integrations

Email me when this repo regresses

Free. We re-scan periodically; new criticals → your inbox. No signup required for the scan itself.

API access

This page is publicly accessible at: https://repobility.com/scan/9e258625-4ffa-414b-bb61-034ab4376b8d/

To check status programmatically (no auth required):

curl -s https://repobility.com/api/v1/public/scan/9e258625-4ffa-414b-bb61-034ab4376b8d/

Important — please don't re-submit the same URL repeatedly. The submission endpoint is idempotent: re-submitting the same git URL returns this same scan_token, not a new one. To re-scan this repo, sign up free and use the dashboard.