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.
Sign up free Scan another repo
77 of your 136 findings came from Repobility's proprietary detections. ✓ Repobility tags below mark them.

Scan timing: clone 6.99s · analysis 5.09s · 10.6 MB · GitHub API rate-limit (preflight)

kubernetes-sigs/inference-perf

https://github.com/kubernetes-sigs/inference-perf · scanned 2026-06-05 19:52 UTC (4 days, 14 hours ago) · 10 languages

234 raw signals (130 security + 104 graph) 56th percentile · Python · medium (20-100K LoC) System graph score 82 (lower by 14)

File as issue Image
UNIFIED Repobility · multi-layer engine · AI coders

Complete repo analysis

Last scanned 4 days, 14 hours ago · v2 · 99 actionable findings from 2 signal sources. 58 repeated signals grouped for readability. Security checks, system graph analysis, and verified AI-agent feedback are merged into one review queue.

JSON
Combined
67.6
/100
Security checks
53
64 findings
System graph
82
100.0% cov · 35 findings
Critical
1
click to filter
Agents
6
0 votes
Crowd
0
reported
Score breakdown â 2026-05-18-v5
Component Sub-score Weight Contribution
structure_score 60.0 0.15 9.00
security_score 32.1 0.25 8.03
testing_score 100.0 0.20 20.00
documentation_score 88.0 0.15 13.20
practices_score 84.0 0.15 12.60
code_quality 47.3 0.10 4.73
Overall 1.00 67.6
Severity distribution — click a segment to filter
Active filters: excluding tests × Reset all
Severity: Critical 1 High 12 Medium 9 Low 25 9-Layer: Software Security Quality Integrity Frontend Hardware Data Network Cicd Source: Security checks 64 System graph 35 Crowd 0 Layer: Software 17 Cicd 8 Security 7 Quality 63 Api 1 Hardware 2 Frontend 1
Scan summary Quality grade B- (68/100). Dimensions: security 32, maintainability 60. 130 findings (13 security). 30,859 lines analyzed.

Showing 45 of 99 actionable findings. 157 raw detector signals were grouped into reader-sized issues. Click TP / FP to vote on a finding's accuracy — votes adjust the confidence weighting and improve detection across the platform.

high Security checks quality Quality conf 1.00 ✓ Repobility 7 occurrences Missing import: `http` used but not imported
The file uses `http.something(...)` but never imports `http`. This raises NameError at runtime the first time the line executes.
5 files, 7 locations
inference_perf/datagen/otel_trace_to_replay_graph.py:1129, 1198 (2 hits)
inference_perf/datagen/replay_graph_session_datagen.py:191, 979 (2 hits)
e2e/utils/llm_d_inference_sim.py:117
inference_perf/loadgen/load_generator.py:378
inference_perf/utils/request_queue.py:67
high Security checks quality Quality conf 1.00 ✓ Repobility [MINED034] Python Subprocess Shell True: subprocess(..., shell=True) enables command injection.
Review and fix per the pattern semantics. See CWE-78 / for context.
scripts/check_coverage_regression.py:36
high Security checks software dependencies conf 0.90 ✓ Repobility 3 occurrences Dockerfile FROM `python:3.12.11-alpine3.22` not pinned by digest
`FROM python:3.12.11-alpine3.22` resolves the tag at build time. The registry CAN re-push a different image for the same tag, so every build is potentially different. Production images should pin to `image@sha256:...` for reproducibility + supply-chain integrity.
2 files, 3 locations
Dockerfile:2, 24 (2 hits)
Dockerfile.e2e-test:1
high Security checks software dependencies conf 0.88 transformers: PYSEC-2025-217
Hugging Face Transformers X-CLIP Checkpoint Conversion Deserialization of Untrusted Data Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of Hugging Face Transformers. User interaction is required to exploit this vul…
pdm.lock
high Security checks software dependencies conf 0.88 urllib3: PYSEC-2026-141
urllib3 is an HTTP client library for Python. From 1.23 to before 2.7.0, cross-origin redirects followed from the low-level API via ProxyManager.connection_from_url().urlopen(..., assert_same_host=False) still forward these sensitive headers. This vulnerability is fixed in 2.7.0.
pdm.lock
high Security checks software dependencies conf 0.88 urllib3: PYSEC-2026-142
urllib3 is an HTTP client library for Python. From 2.6.0 to before 2.7.0, urllib3 could decompress the whole response instead of the requested portion (1) during the second HTTPResponse.read(amt=N) call when the response was decompressed using the official Brotli library or (2) when HTTPResponse.dr…
pdm.lock
high System graph security security conf 1.00 Insecure pattern 'eval_used' in inference_perf/config/datagen/replay.py:128
Found a known-risky pattern (eval_used). Review and replace if possible.
inference_perf/config/datagen/replay.py:128 Eval used
high System graph security security conf 1.00 Insecure pattern 'eval_used' in inference_perf/datagen/otel_trace_replay_datagen.py:228
Found a known-risky pattern (eval_used). Review and replace if possible.
inference_perf/datagen/otel_trace_replay_datagen.py:228 Eval used
low Security checks quality Error handling conf 1.00 [ERR001] Silent Exception Swallowing: Silently swallowing all exceptions hides bugs. Even in cleanup code, log at DEBUG level.
Log the error: `except Exception: logger.debug('cleanup failed', exc_info=True)`. Or handle specific exception types.
inference_perf/circuit_breaker/simple_breaker.py:41
low Security checks security Injection conf 0.50 [SEC005] Command Injection Risk: Unsafe shell execution or eval of user input.
Use subprocess with shell=False and a list of args. Never eval user input.
scripts/check_coverage_regression.py:36
medium Security checks software dependencies conf 0.88 aiohttp: GHSA-hg6j-4rv6-33pg
AIOHTTP is vulnerable to cross-origin redirect with per-request cookies
pdm.lock
medium Security checks software dependencies conf 0.88 aiohttp: GHSA-jg22-mg44-37j8
AIOHTTP is Vulnerable to Deserialization of Untrusted Data
pdm.lock
low Security checks quality Error handling conf 0.55 ✓ Repobility 4 occurrences Broad exception handler needs review
This handler catches Exception/BaseException. It is actionable when it swallows errors without logging, re-raising, or returning a structured error. Handlers that intentionally convert exceptions into typed error results should not be treated as high risk.
3 files, 4 locations
scripts/check_coverage_regression.py:151, 164 (2 hits)
inference_perf/client/modelserver/openai_client.py:399
scripts/check_license.py:45
Error handlingquality
medium Security checks cicd CI/CD security conf 0.94 4 occurrences Compose service `prometheus` image uses the latest tag
The latest tag is mutable and can change without a code review, producing different images from the same source.
3 files, 4 locations
examples/vllm/docker-compose.yml:3, 21 (2 hits)
examples/sglang/docker-compose.yml:24
examples/tgi/docker-compose.yml:21
CI/CD securitycontainers
high Security checks cicd CI/CD security conf 0.82 Docker final stage has no non-root USER
Docker images run as root unless the image or Dockerfile switches to a non-root user.
Dockerfile:24 CI/CD securitycontainers
medium Security checks software dependencies conf 0.88 idna: GHSA-65pc-fj4g-8rjx
Internationalized Domain Names in Applications (IDNA): Specially crafted inputs to idna.encode() can bypass CVE-2024-3651 fix
pdm.lock
medium Security checks software dependencies conf 0.88 transformers: GHSA-69w3-r845-3855
HuggingFace Transformers allows for arbitrary code execution in the `Trainer` class
pdm.lock
medium System graph hardware Security conf 1.00 Dockerfile runs as root: Dockerfile
No non-root USER set. Containers running as root expand the blast radius of any vulnerability inside the image.
Container
medium System graph cicd CI/CD security conf 1.00 GitHub Actions workflow grants broad write permissions
CI tokens with write permissions increase blast radius when an action, dependency, or PR workflow is compromised. Prefer job-level least-privilege permissions.
.github/workflows/publish-on-release.yml CI/CD securitySupply chainGithub actions
medium System graph security security conf 1.00 Insecure pattern 'subprocess_shell_true' in scripts/check_coverage_regression.py:36
Found a known-risky pattern (subprocess_shell_true). Review and replace if possible.
scripts/check_coverage_regression.py:36 Subprocess shell true
medium System graph quality Integrity conf 1.00 Network/subprocess call without timeout or try/except — e2e/utils/testdata.py:41
`subprocess.run(...)` here lacks both a `timeout=` arg and an enclosing try/except. This is exactly the class of bug that took down our git-clone earlier (HTTP/2 stream cancel surfaced as a fatal). Add a `timeout=` and wrap in try/except, or use a wrapper that retries.
runtime safetyRobustness
low Security checks cicd CI/CD security conf 0.72 .dockerignore misses sensitive defaults
.dockerignore exists but does not cover common secret or VCS patterns.
.dockerignore CI/CD securitycontainers
high Security checks cicd CI/CD security conf 0.56 6 occurrences Compose service does not declare a runtime user
If the image does not define USER internally, this service may run as root.
3 files, 6 locations
examples/sglang/docker-compose.yml:3, 24 (2 hits)
examples/tgi/docker-compose.yml:3, 21 (2 hits)
examples/vllm/docker-compose.yml:3, 21 (2 hits)
CI/CD securitycontainers
high Security checks cicd CI/CD security conf 0.62 6 occurrences Compose service lacks no-new-privileges hardening
no-new-privileges prevents processes from gaining additional privileges through setuid binaries or file capabilities.
3 files, 6 locations
examples/sglang/docker-compose.yml:3, 24 (2 hits)
examples/tgi/docker-compose.yml:3, 21 (2 hits)
examples/vllm/docker-compose.yml:3, 21 (2 hits)
CI/CD securitycontainers
low Security checks quality Quality conf 0.60 12 occurrences Duplicated implementation block across source files
Duplicate implementation blocks are maintenance debt. Keep them visible, but they are not a high-severity defect unless the duplicated logic is security-sensitive or drifting.
8 files, 12 locations
inference_perf/client/modelserver/tgi_client.py:9, 10, 131 (3 hits)
inference_perf/client/modelserver/sglang_client.py:9, 191 (2 hits)
inference_perf/client/modelserver/vllm_client.py:9, 23 (2 hits)
inference_perf/datagen/conversation_replay_datagen.py:65
inference_perf/datagen/hf_sharegpt_datagen.py:76
inference_perf/datagen/infinity_instruct_datagen.py:51
inference_perf/datagen/shared_prefix_datagen.py:233
inference_perf/datagen/synthetic_datagen.py:18
duplicationquality
low System graph hardware Supply chain conf 1.00 2 occurrences Docker base image is tag-pinned but not digest-pinned: python:3.12.11-alpine3.22
Container tags can be retagged upstream. Pin production base images to a reviewed digest (`image@sha256:...`) when reproducibility and supply-chain integrity matter.
lines 2, 24
Dockerfile:2, 24 (2 hits)
containersPinned dependencies
low System graph quality Integrity conf 1.00 15 occurrences Near-duplicate function bodies in 2 places
Functions with the same first-5-line body hash: inference_perf/client/server_metrics/base.py:wait, inference_perf/circuit_breaker/base.py:reset This is *the* AI-coder failure mode (4× more duplication in vibe-coded repos — see https://jw.hn/ai-code-hygiene). Consolidate or document why they're sep…
15 occurrences
repo-level (15 hits)
duplicatesduplication
low System graph quality Integrity conf 1.00 Near-duplicate function bodies in 3 places
Functions with the same first-5-line body hash: inference_perf/datagen/synthetic_datagen.py:get_supported_apis, inference_perf/datagen/conversation_replay_datagen.py:get_supported_apis, inference_perf/datagen/random_datagen.py:get_supported_apis This is *the* AI-coder failure mode (4× more duplica…
duplicatesduplication
low System graph quality Integrity conf 1.00 3 occurrences Near-duplicate function bodies in 5 places
Functions with the same first-5-line body hash: inference_perf/client/server_metrics/prometheus_client/base.py:get_queries, inference_perf/client/server_metrics/prometheus_client/base.py:get_queries, inference_perf/client/server_metrics/prometheus_client/base.py:get_queries, inference_perf/client/s…
3 occurrences
repo-level (3 hits)
duplicatesduplication
low System graph quality Integrity conf 1.00 Near-duplicate function bodies in 8 places
Functions with the same first-5-line body hash: inference_perf/datagen/multimodal_datagen.py:is_io_distribution_supported, inference_perf/datagen/synthetic_datagen.py:is_io_distribution_supported, inference_perf/datagen/hf_sharegpt_datagen.py:is_io_distribution_supported, inference_perf/datagen/hf_…
duplicatesduplication
low System graph quality Integrity conf 1.00 Old/deprecated-named symbol `msg_copy` in inference_perf/datagen/replay_graph_session_datagen.py:493
Names with suffixes like `_old`, `_v1`, `_deprecated` usually indicate replaced-but-not-removed code (typical AI-coder leftover). Confirm and delete, or rename if it's the active version.
old markerDead code
low System graph quality Integrity conf 1.00 Old/deprecated-named symbol `recorded_id_old` in tests/test_tool_call_capture.py:317
Names with suffixes like `_old`, `_v1`, `_deprecated` usually indicate replaced-but-not-removed code (typical AI-coder leftover). Confirm and delete, or rename if it's the active version.
old markerDead code
low System graph software Dead code conf 1.00 Possibly dead Python function: adjust
No callers detected by AST scan in this repo. Could be exported for external callers or a framework handler.
inference_perf/datagen/shared_prefix_datagen.py:215
low System graph software Dead code conf 1.00 Possibly dead Python function: adjust_tokens
No callers detected by AST scan in this repo. Could be exported for external callers or a framework handler.
inference_perf/datagen/datagen_utils.py:137
low System graph software Dead code conf 1.00 Possibly dead Python function: adjust_tokens
No callers detected by AST scan in this repo. Could be exported for external callers or a framework handler.
inference_perf/datagen/synthetic_datagen.py:92
low System graph software Dead code conf 1.00 Possibly dead Python function: enrich_metrics
No callers detected by AST scan in this repo. Could be exported for external callers or a framework handler.
inference_perf/metrics/session_collector.py:60
low System graph software Dead code conf 1.00 Possibly dead Python function: generate_span_id
No callers detected by AST scan in this repo. Could be exported for external callers or a framework handler.
inference_perf/client/modelserver/otel_instrumentation.py:85
low System graph software Dead code conf 1.00 Possibly dead Python function: generate_trace_id
No callers detected by AST scan in this repo. Could be exported for external callers or a framework handler.
inference_perf/client/modelserver/otel_instrumentation.py:94
low System graph software Dead code conf 1.00 Possibly dead Python function: new_session
No callers detected by AST scan in this repo. Could be exported for external callers or a framework handler.
inference_perf/client/modelserver/openai_client.py:108
low System graph software Dead code conf 1.00 Possibly dead Python function: new_session
No callers detected by AST scan in this repo. Could be exported for external callers or a framework handler.
inference_perf/client/modelserver/base.py:161
low System graph quality Integrity conf 1.00 Stub function `collect_metrics_summary` (body is just `pass`/`return`) — inference_perf/client/server_metrics/mock_client.py:22
Likely an AI scaffold that was never filled in. Remove or implement.
Empty handlerDead code
low System graph quality Complexity conf 1.00 Very large file: inference_perf/datagen/otel_trace_to_replay_graph.py (1361 lines)
Files with >800 lines often hide complexity hotspots and discourage tests.
low System graph quality Complexity conf 1.00 Very large file: inference_perf/datagen/replay_graph_session_datagen.py (1240 lines)
Files with >800 lines often hide complexity hotspots and discourage tests.
low System graph quality Complexity conf 1.00 Very large file: inference_perf/loadgen/load_generator.py (1167 lines)
Files with >800 lines often hide complexity hotspots and discourage tests.
low System graph quality Complexity conf 1.00 Very large file: tests/test_otel_replay_datagen.py (1504 lines)
Files with >800 lines often hide complexity hotspots and discourage tests.
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/def834f7-1e77-4438-ad9f-7f593101ea91/

To check status programmatically (no auth required): 

curl -s https://repobility.com/api/v1/public/scan/def834f7-1e77-4438-ad9f-7f593101ea91/

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.