AI Slop Code Risk Monitor

deepseek-ai/FlashMLA

SlopCodeMonitor badge for deepseek-ai/FlashMLA Adjusted legacy signal; badge bands: 25 / 50 / 75 (different from dashboard bands).

FlashMLA: Efficient Multi-head Latent Attention Kernels

4.5 Low pattern signal View on GitHub

Open-Source Code Quality, Security and Technical Debt Review

This report summarizes scanner evidence for deepseek-ai/FlashMLA, a C++ project with 13,225 GitHub stars. SynthScan examined 21,503 lines of code across 90 source files, recording 102 pattern matches distributed across 7 syntactic categories. The legacy adjusted pattern signal is 4.5 (Low pattern signal); it is not an authorship estimate.

Use the findings to ask whether maintainers can explain intent, verify dependencies, simplify unnecessary scaffolding, and focus review effort. Risk Payload v1 adds ownership, supply-chain, over-engineering, and review-effort vectors when available. These are prompts for human verification, not probabilities, security findings, or measured maintenance costs.

4.5
Adjusted Signal
4.5
Raw Pattern Score
100%
Time Factor
2026-10-27
Last Push
13.2K
Stars
C++
Language
21.5K
Lines of Code
90
Files
102
Pattern Hits
2026-10-31
Scan Date
0.00
HC Hit Rate

Risk Review · Payload v1

Use each vector to guide manual verification of the underlying findings.

Payload schema: 1.0.0 · Engine: 2.0.0 · Scoring: heuristic-density-v1 · Scan date: 2026-10-31.

Cognitive Ownership Risk 11.1/100 · LOW

Can maintainers explain intent, assumptions, and trade-offs? 49 finding(s).

Hallucinated Dependency & Supply Chain Risk 9.3/100 · LOW

Are imports declared and package/API names verified? 49 finding(s).

Agent Over-Engineering Ratio 6.8/100 · LOW

Are wrappers, guards, and repeated structures necessary? 28 finding(s).

PR Review Velocity Impact 21.5/100 · LOW

Which changes need focused human review? 102 finding(s).

What These Metrics Mean

Adjusted Signal
Legacy pattern score normalized per 1,000 scanned lines and multiplied by the repository-level time factor. It is not an estimate of AI authorship.
Raw Pattern Score
Sum of severity- and context-weighted pattern findings before applying the time factor.
Time Factor
A coarse multiplier based on repository last-push metadata. It does not identify when individual code was written.
Pattern Hits
Count of configured pattern findings. Review each location and surrounding code before acting.
HC Hit Rate
High- and critical-severity findings per scanned file. Severity is a configured detector weight, not proof of a vulnerability or authorship.
Lines / Files
Scan coverage denominators used by the legacy normalized score.

Pattern-Signal History

This chart tracks the legacy adjusted pattern score across scan runs. Changes can reflect source changes, scan coverage, or detector/scoring versions; they do not measure AI adoption or prove that a team refactored code. Compare runs only when their scoring versions are compatible. The dashed line tracks the raw pattern-hit count.

Severity Breakdown

Groups findings by configured severity weight. Severity is not a confidence estimate, security rating, or proof that a package is missing. Verify imports against the project's manifests and inspect each finding in its local design context.

CRITICAL 0HIGH 0MEDIUM 1LOW 101

Directory Score Breakdown

This horizontal bar chart shows where configured pattern scores accumulate across directories. Treat the bars as locations to inspect, not as evidence of authorship or code quality. Use the file-level findings to decide whether a directory needs a focused ownership, dependency, complexity, or review check.

Pattern Findings

The scanner identified 102 distinct pattern matches across 7 syntactic categories. Each entry below represents a discrete location in the source code where the engine recorded a detector match to inspect in context. Expand any category row to inspect the individual file paths, line numbers, code snippets, and the lexical context (CODE, COMMENT, or STRING) in which each match was detected.

Reading the findings table: The Severity column indicates the configured finding weight (CRITICAL / HIGH / MEDIUM / LOW). The Context column identifies whether the match occurred inside executable code, an inline comment, or a string literal — legacy comment-context matches receive a ×1.5 score weight. The ⚡ bolt icon marks clustered matches: three or more patterns within a 10-line window, each receiving an additional ×1.5 density multiplier under the legacy scoring model.

Unused Imports49 hits · 42 pts
SeverityFileLineSnippetContext
LOWtests/lib.py11CODE
LOWtests/kernelkit/__init__.py1CODE
LOWtests/kernelkit/__init__.py2CODE
LOWtests/kernelkit/__init__.py3CODE
LOWtests/kernelkit/__init__.py4CODE
LOWtests/kernelkit/__init__.py5CODE
LOWtests/kernelkit/__init__.py6CODE
LOWtests/kernelkit/__init__.py7CODE
LOWtests/kernelkit/__init__.py9CODE
LOWtests/kernelkit/__init__.py9CODE
LOWtests/kernelkit/__init__.py10CODE
LOWtests/kernelkit/__init__.py10CODE
LOWtests/kernelkit/__init__.py11CODE
LOWtests/kernelkit/__init__.py11CODE
LOWtests/kernelkit/__init__.py11CODE
LOWtests/kernelkit/__init__.py11CODE
LOWtests/kernelkit/__init__.py11CODE
LOWtests/kernelkit/__init__.py12CODE
LOWtests/kernelkit/__init__.py12CODE
LOWtests/kernelkit/__init__.py12CODE
LOWtests/kernelkit/__init__.py13CODE
LOWtests/kernelkit/__init__.py13CODE
LOWtests/kernelkit/__init__.py13CODE
LOWtests/kernelkit/__init__.py13CODE
LOWtests/kernelkit/__init__.py13CODE
LOWtests/kernelkit/__init__.py13CODE
LOWtests/kernelkit/__init__.py13CODE
LOWtests/kernelkit/__init__.py14CODE
LOWtests/kernelkit/__init__.py14CODE
LOWtests/kernelkit/__init__.py14CODE
LOWtests/kernelkit/__init__.py15CODE
LOWtests/kernelkit/__init__.py15CODE
LOWtests/kernelkit/__init__.py15CODE
LOWtests/kernelkit/__init__.py15CODE
LOWtests/kernelkit/__init__.py15CODE
LOWtests/kernelkit/bench.py1CODE
LOWtests/kernelkit/utils.py5CODE
LOWtests/kernelkit/utils.py5CODE
LOWtests/kernelkit/compare.py1CODE
LOWtests/kernelkit/compare.py1CODE
LOWtests/kernelkit/compare.py1CODE
LOWflash_mla/__init__.py34CODE
LOWflash_mla/__init__.py34CODE
LOWflash_mla/__init__.py34CODE
LOWflash_mla/__init__.py34CODE
LOWflash_mla/__init__.py34CODE
LOWflash_mla/__init__.py34CODE
LOWflash_mla/__init__.py43CODE
LOWflash_mla/__init__.py19CODE
Hyper-Verbose Identifiers21 hits · 22 pts
SeverityFileLineSnippetContext
LOWsetup.py324def get_build_target_platform() -> str:CODE
LOWtests/test-fused-norm-rope-attn-rope-cast.py83def ref_fused_norm_rope_attn_rope_cast_fwd(p: TestParam, t: Testcase, enable_q_norm: bool, cos_sin_table: torch.Tensor, CODE
LOWtests/test-fused-norm-rope-attn-rope-cast.py107def ref_fused_norm_rope_attn_rope_cast_decode(CODE
LOWtests/test-fused-norm-rope-attn-rope-cast.py238 def run_fused_norm_rope_attn_rope_cast_fwd():CODE
LOWtests/test-fused-norm-rope-attn-rope-cast.py335 def run_fused_norm_rope_attn_rope_cast_decode():CODE
LOWtests/lib.py20def get_current_compute_capability() -> Tuple[int, int]:CODE
LOWtests/lib.py268 def get_kvcache_for_flash_mla(self) -> torch.Tensor:CODE
LOWtests/lib.py284def generate_testcase_for_decode(t: TestParam) -> TestcaseForDecode:CODE
LOWtests/lib.py463def count_flop_and_mem_vol_for_decode(p: TestParam, t: TestcaseForDecode) -> FlopsAndMemVolStatisticsForDecode:CODE
LOWtests/kernelkit/build.py17def check_kernel_reg_spill_in_artifact(artifact_path: str, stack_baseline: int = 0, quiet: bool = False, suppress_checkiCODE
LOW⚡tests/kernelkit/generate.py3def _get_new_non_contiguous_tensor_shape(shape):CODE
LOW⚡tests/kernelkit/generate.py10def gen_non_contiguous_randn_tensor(shape, *args, **kwargs):CODE
LOW⚡tests/kernelkit/generate.py16def gen_non_contiguous_tensor(shape, *args, **kwargs):CODE
LOWtests/kernelkit/bench.py24def profiler_range_start_marker_kernel():CODE
LOWtests/kernelkit/bench.py27def _run_profiler_range_start_marker_kernel():CODE
LOWtests/kernelkit/compare.py7def check_is_bitwise_equal_comparator(ans: torch.Tensor, ref: torch.Tensor, result: torch.Tensor):CODE
LOWtests/kernelkit/compare.py107def check_is_allclose_comparator(name: str, ans: torch.Tensor, ref: torch.Tensor, out: torch.Tensor, abs_tol: float = 1eCODE
LOWflash_mla/flash_mla_interface.py235def _flash_attn_varlen_forward(CODE
LOWflash_mla/flash_mla_interface.py285def _flash_attn_varlen_backward(CODE
LOWflash_mla/flash_mla_interface.py458def flash_attn_varlen_qkvpacked_func(CODE
LOWflash_mla/flash_mla_interface.py505def flash_attn_varlen_kvpacked_func(CODE
Over-Commented Block16 hits · 16 pts
SeverityFileLineSnippetContext
LOWcsrc/params.h1#pragma onceCOMMENT
LOWcsrc/params.h21 struct bf16;COMMENT
LOWcsrc/device_stream.h1#pragma onceCOMMENT
LOWcsrc/api/common.h1#pragma onceCOMMENT
LOWcsrc/api/common.h21#include <torch/headeronly/util/shim_utils.h>COMMENT
LOWcsrc/api/ascend_stream.h1#pragma onceCOMMENT
LOWcsrc/api/ascend_stream.cpp1#include "ascend_stream.h"COMMENT
LOW…100/prefill/sparse/fwd_for_small_topk/head128/config.h1#pragma onceCOMMENT
LOWcsrc/3rdparty/kerutils/include/kerutils/host/host.h1#pragma onceCOMMENT
LOW…kerutils/include/kerutils/supplemental/torch_tensors.h61#ifdef KERUTILS_IS_BUILD_ON_CUDACOMMENT
LOW…kerutils/include/kerutils/supplemental/torch_tensors.h81// Check whether the given tensor (or optional<tensor>) has the specified dtypeCOMMENT
LOWcsrc/3rdparty/kerutils/include/kerutils/common/common.h1#pragma onceCOMMENT
LOW…rdparty/kerutils/include/kerutils/device/cuda/common.h41COMMENT
LOW…rdparty/kerutils/include/kerutils/device/cuda/common.h61#define KERUTILS_ENABLE_SM100ACOMMENT
LOWcsrc/ascend_kernels/prefill/sparse/config.h101 int32_t indices[INDEX_BLOCK_SIZE];COMMENT
LOWtests/__init__.py1# We use this `__init__.py` to let `setup.py` be able to import `tests.kernelkit`COMMENT
Deep Nesting9 hits · 9 pts
SeverityFileLineSnippetContext
LOWsetup.py137CODE
LOWsetup.py160CODE
LOWtests/lib.py68CODE
LOWtests/quant.py58CODE
LOWtests/quant.py139CODE
LOWtests/kernelkit/bench.py112CODE
LOWtests/kernelkit/bench.py177CODE
LOWtests/kernelkit/bench.py264CODE
LOWtests/kernelkit/bench.py51CODE
Excessive Try-Catch Wrapping3 hits · 4 pts
SeverityFileLineSnippetContext
LOWsetup.py349 except Exception:CODE
MEDIUMsetup.py344def get_version() -> str:CODE
LOWtests/kernelkit/bench.py279 except Exception as e:CODE
AI Structural Patterns3 hits · 3 pts
SeverityFileLineSnippetContext
LOWtests/quant.py194CODE
LOWtests/kernelkit/bench.py110CODE
LOWflash_mla/flash_mla_interface.py55CODE
Modern Structural Boilerplate1 hit · 1 pts
SeverityFileLineSnippetContext
LOWflash_mla/__init__.py45__all__ = [CODE