LMCache/LMCache

37.4

Adjusted Score

37.4

Raw Score

100%

Time Factor

2026-08-28

Last Push

11.6K

Stars

Python

Language

490.1K

Lines of Code

1.9K

Files

10.8K

Pattern Hits

2026-08-29

Scan Date

0.24

HC Hit Rate

What These Metrics Mean

Adjusted Score: Primary synthetic code indicator. Raw score normalised per 1,000 lines of code and multiplied by the temporal discount factor. This is the definitive comparative metric — use it to rank repositories by AI authorship density.
Raw Score: The unmodified sum of all severity-weighted, context-multiplied pattern match scores before temporal discounting. Reflects the absolute signal strength independent of when the repository was last active.
Time Factor: The temporal discount multiplier (0–100%) applied to the raw score. Repositories last updated before ChatGPT's launch (Nov 2022) receive a 5% factor. Full signal is only assigned to repositories active in the post-adoption era (Jan 2024+).
Pattern Hits: Total count of individual pattern matches across all files and categories. A high hit count with a low score may indicate a very large codebase with isolated AI snippets; a low count with a high score indicates dense, concentrated AI signatures.
HC Hit Rate: High+Critical pattern hits per file, averaged across the repository. This orthogonal signal catches repositories where a few files are densely packed with high-severity AI tells — a strong indicator even when the normalised score appears moderate due to codebase size.
Lines of Code / Files: Total lines and files analysed. The scanner examines 94 file extensions. These denominators are used to normalise the score, enabling fair comparison between repositories of vastly different sizes.

Score History

This chart maps the temporal evolution of the adjusted synthetic code score across successive scan runs. An upward trajectory indicates ongoing incorporation of AI-generated code or expanding LLM-assisted scaffolding; a stable or declining trajectory may reflect active human refactoring, code removal, or the adoption of stricter authorship policies. The dashed secondary line (right axis) independently tracks total raw pattern hit count, which can diverge from the normalised score when codebase size changes significantly between scans.

Severity Breakdown

Classifies detected patterns by their diagnostic confidence and structural impact. CRITICAL patterns (coefficient 10) represent definitive synthetic signatures — hallucinated imports, explicit LLM attribution metadata — virtually never produced by human authors. HIGH (5) indicates strong structural tells such as cross-file repetition or cross-linguistic idioms. MEDIUM (2) covers recognisable conversational padding and AI-specific vocabulary. LOW (1) captures subtle indicators like tautological comments and generic boilerplate that require density to carry independent signal.

CRITICAL 2HIGH 457MEDIUM 2238LOW 8059

Directory Score Breakdown

This horizontal bar chart decomposes the repository's raw synthetic code score by top-level directory, allowing you to pinpoint precisely which modules or components carry the highest AI authorship density. Directories with disproportionately high scores relative to their size warrant targeted manual review: concentrated AI signatures often trace back to mass-generated configuration layers, auto-ported test suites, LLM-scaffolded boilerplate classes, or entire subsystems authored under heavy copilot assistance. Use this view to prioritise your human code-review effort.

Pattern Findings

The scanner identified 10756 distinct pattern matches across 24 syntactic categories. Each entry below represents a discrete location in the source code where the engine recorded a statistically significant AI authorship indicator. 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 diagnostic confidence level (CRITICAL / HIGH / MEDIUM / LOW). The Context column identifies whether the match occurred inside executable code, an inline comment, or a string literal — comment-context matches receive a ×1.5 weight because LLMs systematically over-annotate. The ⚡ bolt icon marks clustered matches: three or more patterns within a 10-line window, each receiving an additional ×1.5 density multiplier as dense clusters constitute far stronger evidence of synthetic authorship than isolated hits.

Decorative Section Separators2059 hits · 7550 pts

Severity	File	Line	Snippet	Context
MEDIUM⚡	tests/test_utils.py	496	# ============================================================	COMMENT
MEDIUM⚡	tests/test_utils.py	498	# ============================================================	COMMENT
MEDIUM	tests/test_utils.py	27	# ============================================================	COMMENT
MEDIUM	tests/test_utils.py	29	# ============================================================	COMMENT
MEDIUM	tests/test_utils.py	61	# ============================================================	COMMENT
MEDIUM	tests/test_utils.py	63	# ============================================================	COMMENT
MEDIUM	tests/test_utils.py	188	# ============================================================	COMMENT
MEDIUM	tests/test_utils.py	190	# ============================================================	COMMENT
MEDIUM	tests/test_utils.py	211	# ============================================================	COMMENT
MEDIUM	tests/test_utils.py	213	# ============================================================	COMMENT
MEDIUM	tests/test_utils.py	234	# ============================================================	COMMENT
MEDIUM	tests/test_utils.py	236	# ============================================================	COMMENT
MEDIUM	tests/test_utils.py	394	# ============================================================	COMMENT
MEDIUM	tests/test_utils.py	396	# ============================================================	COMMENT
MEDIUM⚡	tests/v1/test_health_monitor_fallback_recovery.py	186	# ==============================================================================	COMMENT
MEDIUM⚡	tests/v1/test_health_monitor_fallback_recovery.py	188	# ==============================================================================	COMMENT
MEDIUM	tests/v1/test_health_monitor_fallback_recovery.py	94	# ==============================================================================	COMMENT
MEDIUM	tests/v1/test_health_monitor_fallback_recovery.py	96	# ==============================================================================	COMMENT
MEDIUM⚡	tests/v1/test_torch_ops.py	124	# ==========================================	COMMENT
MEDIUM⚡	tests/v1/test_torch_ops.py	126	# ==========================================	COMMENT
MEDIUM	tests/v1/test_torch_ops.py	22	# ==========================================	COMMENT
MEDIUM	tests/v1/test_torch_ops.py	24	# ==========================================	COMMENT
MEDIUM	tests/v1/test_torch_ops.py	50	# ==========================================	COMMENT
MEDIUM	tests/v1/test_torch_ops.py	52	# ==========================================	COMMENT
MEDIUM	tests/v1/test_torch_ops.py	2891	# ==========================================	COMMENT
MEDIUM	tests/v1/test_torch_ops.py	2893	# ==========================================	COMMENT
MEDIUM	tests/v1/test_torch_ops.py	2922	# ==========================================	COMMENT
MEDIUM	tests/v1/test_torch_ops.py	2924	# ==========================================	COMMENT
MEDIUM	tests/v1/test_torch_ops.py	3020	# ==========================================	COMMENT
MEDIUM	tests/v1/test_torch_ops.py	3022	# ==========================================	COMMENT
MEDIUM	tests/v1/test_musa_support.py	38	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_musa_support.py	40	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/v1/test_musa_support.py	89	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/v1/test_musa_support.py	91	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/v1/test_musa_support.py	169	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/v1/test_musa_support.py	171	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/v1/test_musa_support.py	303	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/v1/test_musa_support.py	305	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_nixl_batched_contains.py	27	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_nixl_batched_contains.py	29	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_nixl_batched_contains.py	73	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_nixl_batched_contains.py	75	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_nixl_batched_contains.py	167	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_nixl_batched_contains.py	169	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_nixl_batched_contains.py	226	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_nixl_batched_contains.py	228	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_device_id_race.py	22	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_device_id_race.py	25	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/v1/test_lazy_offload_pending_store.py	26	# ===========================================================================	COMMENT
MEDIUM⚡	tests/v1/test_lazy_offload_pending_store.py	28	# ===========================================================================	COMMENT
MEDIUM⚡	tests/v1/test_lazy_offload_pending_store.py	126	# ===========================================================================	COMMENT
MEDIUM⚡	tests/v1/test_lazy_offload_pending_store.py	128	# ===========================================================================	COMMENT
MEDIUM⚡	tests/v1/test_mp_mem_kernels.py	30	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/v1/test_mp_mem_kernels.py	32	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_mp_mem_kernels.py	235	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_mp_mem_kernels.py	237	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_mp_mem_kernels.py	284	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_mp_mem_kernels.py	286	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_mp_mem_kernels.py	517	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/v1/test_mp_mem_kernels.py	525	# ---------------------------------------------------------------------------	COMMENT
1999 more matches not shown…

Hyper-Verbose Identifiers5428 hits · 5317 pts

Severity	File	Line	Snippet	Context
LOW	tools/translate_docs_zh.py	257	def clean_translation_response(text: str) -> str:	CODE
LOW⚡	tests/test_utils.py	156	def test_nested_brackets_error(self):	CODE
LOW⚡	tests/test_utils.py	162	def test_unmatched_closing_bracket(self):	CODE
LOW⚡	tests/test_utils.py	172	def test_invalid_range_start_gt_end(self):	CODE
LOW⚡	tests/test_utils.py	297	def test_from_string_roundtrip(self, basic_key):	CODE
LOW⚡	tests/test_utils.py	302	def test_from_string_with_tags_roundtrip(self, key_with_tags):	CODE
LOW⚡	tests/test_utils.py	311	def test_from_string_invalid_tag_format(self):	CODE
LOW⚡	tests/test_utils.py	315	def test_to_dict_and_from_dict(self, basic_key):	CODE
LOW⚡	tests/test_utils.py	325	def test_from_dict_invalid_tag(self):	CODE
LOW⚡	tests/test_utils.py	500	def test_parse_cache_engine_key(self):	CODE
LOW	tests/test_utils.py	40	def test_numerator_smaller_than_denominator(self):	CODE
LOW	tests/test_utils.py	75	def test_three_or_more_consecutive(self):	CODE
LOW	tests/test_utils.py	83	def test_non_consecutive_order_preserved(self):	CODE
LOW	tests/test_utils.py	109	def test_three_or_more_elements(self):	CODE
LOW	tests/test_utils.py	273	def test_not_equal_different_hash(self, basic_key):	CODE
LOW	tests/test_utils.py	283	def test_not_equal_different_type(self, basic_key):	CODE
LOW	tests/test_utils.py	360	def test_chunk_hash_hex_bytes(self):	CODE
LOW	tests/test_utils.py	372	def test_unsupported_dtype_raises(self):	CODE
LOW	tests/test_utils.py	382	def test_no_tags_when_no_tag_prefix(self):	CODE
LOW	tests/test_utils.py	409	def test_hash_includes_layer_id(self, layer_key):	CODE
LOW	tests/test_utils.py	432	def test_eq_different_base_key(self, layer_key):	CODE
LOW	tests/test_utils.py	462	def test_from_string_roundtrip(self, layer_key):	CODE
LOW	tests/test_utils.py	467	def test_from_string_with_tags(self):	CODE
LOW	tests/test_utils.py	485	def test_from_string_invalid_tag(self):	CODE
LOW	tests/test_utils.py	516	def test_parse_layer_cache_engine_key(self):	CODE
LOW	tests/test_utils.py	534	def test_parse_cache_key_without_tags(self):	CODE
LOW	tests/test_utils.py	549	def test_parse_cache_key_with_tags(self):	CODE
LOW	tests/test_utils.py	568	def test_parse_layer_cache_key_without_tags(self):	CODE
LOW	tests/test_utils.py	583	def test_parse_layer_cache_key_with_tags(self):	CODE
LOW	tests/test_utils.py	601	def test_roundtrip_cache_engine_key_without_tags(self):	CODE
LOW	tests/test_utils.py	621	def test_roundtrip_cache_engine_key_with_tags(self):	CODE
LOW	tests/test_utils.py	644	def test_roundtrip_layer_cache_engine_key_without_tags(self):	CODE
LOW	tests/test_utils.py	665	def test_roundtrip_layer_cache_engine_key_with_tags(self):	CODE
LOW	tests/test_observability.py	32	def test_on_retrieve_finished(stats_monitor):	CODE
LOW⚡	tests/test_observability.py	44	def test_on_store_request_and_finished(stats_monitor):	CODE
LOW⚡	tests/test_observability.py	52	def test_update_local_cache_usage(stats_monitor):	CODE
LOW⚡	tests/test_observability.py	58	def test_update_remote_cache_usage(stats_monitor):	CODE
LOW⚡	tests/test_observability.py	64	def test_update_local_storage_usage(stats_monitor):	CODE
LOW	tests/test_observability.py	104	def test_remote_write_metrics(stats_monitor):	CODE
LOW	tests/test_observability.py	143	def test_retrieve_and_store_speed(stats_monitor):	CODE
LOW	tests/test_observability.py	161	def test_multiple_lookup_operations(stats_monitor):	CODE
LOW	tests/test_observability.py	180	def test_mixed_remote_operations(stats_monitor):	CODE
LOW	tests/test_observability.py	273	def test_zero_division_protection(stats_monitor):	CODE
LOW	tests/test_observability.py	284	def _cleanup_prometheus_logger():	CODE
LOW	tests/test_observability.py	318	def test_prometheus_logger_default_buckets(_cleanup_prometheus_logger):	CODE
LOW	tests/test_observability.py	326	def test_prometheus_logger_custom_buckets(_cleanup_prometheus_logger):	CODE
LOW	tests/test_observability.py	346	def test_prometheus_logger_partial_custom_buckets(	CODE
LOW	tests/test_observability.py	373	def test_prometheus_logger_get_or_create_with_config(	CODE
LOW	tests/test_observability.py	393	def test_prometheus_logger_get_or_create_allows_multiple_roles(	CODE
LOW	tests/test_observability.py	408	def test_prometheus_logger_label_view_rebinds_dynamic_metrics(	CODE
LOW	tests/test_observability.py	438	def test_prometheus_logger_reset_reinitializes_all_label_views(	CODE
LOW	tests/test_banner.py	35	def test_banner_contains_version_links_and_opt_out_hint():	CODE
LOW⚡	tests/test_banner.py	46	def test_banner_is_plain_on_non_tty_stream():	CODE
LOW⚡	tests/test_banner.py	52	def test_banner_is_colored_on_tty_stream():	CODE
LOW⚡	tests/test_banner.py	60	def test_banner_prints_at_most_once_per_process():	CODE
LOW⚡	tests/test_banner.py	70	def test_banner_disabled_by_env_var(monkeypatch, value):	CODE
LOW⚡	tests/test_banner.py	77	def test_banner_not_disabled_by_falsy_env_var(monkeypatch):	CODE
LOW	tests/test_build_profiles_rocm.py	64	def test_hipify_wrapper_returns_paths_relative_to_project_root(	CODE
LOW	tests/test_build_profiles_rocm.py	81	def test_hipify_wrapper_copies_headers_into_the_output_tree(	CODE
LOW	tests/test_serde.py	102	def test_cachegen_unmatched_size():	CODE
5368 more matches not shown…

Docstring Block Structure274 hits · 1370 pts

Severity	File	Line	Snippet	Context
HIGH	tests/v1/distributed/test_memory_capacity.py	122	Build a config whose derived capacity equals ``capacities``. Args: capacities: The per-medium result the co	STRING
HIGH	setup_extensions/build_profiles/rocm.py	28	Hipify ``csrc/cuda`` and return paths for the requested source files. Args: source_names: Base names of the	STRING
HIGH	examples/token_dropping/utils.py	163	Move key cache RoPE from old positions to new positions. Args: kv_tensor: The KV cache tensor of shape [2, L	STRING
HIGH	examples/token_dropping/utils.py	234	Score generated responses against ground-truth choice letters. Args: ground_truth: list of ground-truth let	STRING
HIGH	benchmarks/long_doc_qa/long_doc_qa.py	333	Repeat each prompt in the list for a specified number of times. The order of prompts in the output list depends	STRING
HIGH	…dkite/k3_tests/multiprocess/scripts/run-hma-lm-eval.sh	184	Return the gsm8k (exact_match, stderr) from an lm_eval results directory. Prefers the strict-match variant; falls b	STRING
HIGH	lmcache/v1/memory_management.py	1389	Allocate a block of memory from the virtual address space of a given size. The actual allocated size co	STRING
HIGH	lmcache/v1/memory_management.py	1438	Allocate blocks of memory from the virtual address space of a given size and batch size. The actual all	STRING
HIGH	lmcache/v1/hidden_state_store.py	123	Store hidden_states chunked under the same keys as KV. Args: token_ids: 1-D int tensor or list of t	STRING
HIGH	lmcache/v1/kv_layer_groups.py	82	Partition layer indices by :data:`LayerGroupIdentity`. Args: kv_caches: Registered KV cache structure, one	STRING
HIGH	lmcache/v1/kv_layer_groups.py	644	Calculate the number of blocks for a given number of tokens in a specified kernel group. Args:	STRING
HIGH	lmcache/v1/protocol.py	104	Pad a shape with fewer than 4 dimensions to 4D using trailing zeros. Shapes that are already 4D are returned as	STRING
HIGH	lmcache/v1/config_base.py	51	Parse local disk path configuration. Accepts a single path or comma-separated paths, each optionally prefixed w	STRING
HIGH	lmcache/v1/rpc_utils.py	185	Keep an IPC socket path within the ``sockaddr_un`` length limit. The descriptive path is preferred for readability/	STRING
HIGH	lmcache/v1/memory_allocators/mixed_memory_allocator.py	98	Allocate one object from the mixed pinned/buffer allocator. Args: shapes: Logical tensor shape or s	STRING
HIGH	lmcache/v1/memory_allocators/mixed_memory_allocator.py	139	Allocate multiple objects from the mixed pinned/buffer allocator. Args: shapes: Logical tensor shap	STRING
HIGH	…v1/memory_allocators/paged_cpu_gpu_memory_allocator.py	99	Allocate from the selected CPU or GPU paged allocator. Args: shapes: Logical tensor shape or shapes	STRING
HIGH	…v1/memory_allocators/paged_cpu_gpu_memory_allocator.py	128	Allocate a batch from the selected CPU or GPU paged allocator. Args: shapes: Logical tensor shape o	STRING
HIGH	lmcache/v1/memory_allocators/lazy_memory_allocator.py	28	Get the NUMA ID for the current GPU Args: numa_mapping (NUMAMapping): The NUMA mapping object. Re	STRING
HIGH	lmcache/v1/memory_allocators/devdax_memory_allocator.py	274	Map one Device-DAX device and append it to the pool as ACTIVE. The caller must hold ``host_mem_lock`` (except d	STRING
HIGH	lmcache/v1/memory_allocators/devdax_memory_allocator.py	709	Map an additional Device-DAX device and add it to the pool. The new arena is immediately available as overflow;	STRING
HIGH	lmcache/v1/memory_allocators/devdax_memory_allocator.py	753	Retire a Device-DAX arena from the pool. The arena stops accepting new allocations. If it has no live allocatio	STRING
HIGH	lmcache/v1/mp_coordinator/discovery.py	61	Return the instance of ``member_type``, building it once. Args: member_type: The class to look up.	STRING
HIGH	lmcache/v1/mp_coordinator/schemas.py	50	Decode a base64 token string produced by :func:`encode_tokens`. Args: tokens_b64: Base64 of a little-endian	STRING
HIGH	lmcache/v1/mp_coordinator/schemas.py	297	Enforce encoding-level constraints on every entry. Args: value: The hydrated batches. Retu	STRING
HIGH	lmcache/v1/mp_coordinator/schemas.py	361	Reject undecodable keys at request validation (surfaced as 422). Args: value: The hydrated keys.	STRING
HIGH	lmcache/v1/mp_coordinator/schemas.py	465	Reject a malformed token buffer at request validation. Validating here returns 422 rather than the 500 a decode	STRING
HIGH	lmcache/v1/mp_coordinator/blend_client.py	161	Build a started client for ``url``; an empty URL returns ``None``. Args: url: Coordinator base URL;	STRING
HIGH	lmcache/v1/mp_coordinator/registrar.py	45	Register an MP server with the coordinator and return its id. Args: client: The HTTP client to send with.	STRING
HIGH	lmcache/v1/mp_coordinator/utils/encoding.py	32	Rebuild a key from :func:`encode_key`. Args: encoded: An :func:`encode_key` value, as read back from an	STRING
HIGH	lmcache/v1/mp_coordinator/http_apis/metrics_api.py	16	Return Prometheus metrics when the coordinator uses pull mode. Args: request: Incoming request carrying the	STRING
HIGH	…che/v1/mp_coordinator/http_apis/instances_usage_api.py	197	Return one MP server's memory status. Args: instance_id: The server to report on. request: Carries	STRING
HIGH	lmcache/v1/mp_coordinator/http_apis/directory_api.py	42	Resolve keys — or a request's token sequence — to placements. The tokens form hashes ``token_ids`` with the fleet's	STRING
HIGH	lmcache/v1/mp_coordinator/http_apis/cache_api.py	50	Submit a warm prefetch of a token sequence on one MP server. Resolves ``body.instance_id`` in the registry and prox	STRING
HIGH	lmcache/v1/mp_coordinator/http_apis/cache_api.py	104	Proxy a warm-prefetch status poll to the owning MP server. The warm holds no lock, so this poll only reports progre	STRING
HIGH	lmcache/v1/mp_coordinator/http_apis/cache_api.py	151	Pin a token sequence's keys in the L2 eviction plan. The coordinator resolves the token sequence to its object keys	STRING
HIGH	lmcache/v1/mp_coordinator/http_apis/cache_api.py	190	Unpin a token sequence's keys from the L2 eviction plan. Symmetric with pin: resolves the keys locally and releases	STRING
HIGH	lmcache/v1/mp_coordinator/http_apis/cache_api.py	231	Delete a token sequence on one MP server, per ``body.tier`` (l1 / l2 / all). Args: body: Target instance, m	STRING
HIGH	lmcache/v1/mp_coordinator/http_apis/dependencies.py	63	Return the per-app :class:`CoordinatorContext`. Args: request: The FastAPI request whose ``app.state`` carr	STRING
HIGH	lmcache/v1/mp_coordinator/http_apis/dependencies.py	82	Return the lifespan-bound outbound client, or raise if unset. Args: request: The FastAPI request. Retu	STRING
HIGH	lmcache/v1/mp_coordinator/persistence/metadata.py	127	Parse a document and check its shape. Args: raw: The stored bytes. Returns: Th	STRING
HIGH	lmcache/v1/mp_coordinator/persistence/checkpoint.py	151	Decode a checkpoint written by :func:`_write_checkpoint`. Args: stream: Positioned at the start of the chec	STRING
HIGH	lmcache/v1/mp_coordinator/persistence/quiesce.py	77	Hold while capturing; no batch is applied for the duration. Args: timeout: Seconds to wait for in-f	STRING
HIGH	…ache/v1/mp_coordinator/controllers/prefetch_manager.py	45	``POST /cache/prefetches`` to ``target`` and return its JSON reply. Args: target: The MP server to	STRING
HIGH	…ache/v1/mp_coordinator/controllers/prefetch_manager.py	82	Proxy ``GET /cache/prefetches/{request_id}`` on ``target``. Args: target: The MP server holding the	STRING
HIGH	lmcache/v1/mp_coordinator/views/key_directory.py	291	List keys whose placements match the filters, one page at a time. A snapshot for inspection: iteration order is	STRING
HIGH	lmcache/v1/multiprocess/server.py	181	Assemble the list of engine modules based on configuration. Args: ctx: The shared engine context. m	STRING
HIGH	lmcache/v1/multiprocess/config.py	698	Convert parsed command line arguments to a CoordinatorConfig. For the registration settings a flag value takes prec	STRING
HIGH	lmcache/v1/multiprocess/engine_context.py	162	Look up the attention-window descriptor for a pair. Args: model_name: The model name. w	STRING
HIGH	lmcache/v1/multiprocess/engine_context.py	282	Resolve per-object-group object keys from an IPC cache key. Uses the session manager to track token state and t	STRING
HIGH	lmcache/v1/multiprocess/protocol.py	33	Get the expected payload classes for a request type. Args: req_type: The request type to look up	STRING
HIGH	lmcache/v1/multiprocess/protocol.py	52	Get the expected response class for a request type. Args: req_type: The request type to look up R	STRING
HIGH	lmcache/v1/multiprocess/protocol.py	71	Get the handler type for a request type. Args: req_type: The request type to look up Returns:	STRING
HIGH	lmcache/v1/multiprocess/session.py	70	Compute and return chunk hashes for the [start, end) token range. Internally computes rolling hashes up to end_	STRING
HIGH	lmcache/v1/multiprocess/futures.py	44	Get the result of the future. Args: timeout (Optional[float]): Maximum time to wait in sec	STRING
HIGH	lmcache/v1/multiprocess/futures.py	159	Wait for the future to be done, ordered through the device event. Args: timeout (Optional[	STRING
HIGH	lmcache/v1/multiprocess/futures.py	191	Get the result of the future. Args: timeout (Optional[float]): Maximum time to wait for th	STRING
HIGH	lmcache/v1/multiprocess/posix_shm.py	123	Create a new shared-memory segment and return its mmap address. Mirrors the previous ``shm_open(O_CREAT\|O_EXCL) + f	STRING
HIGH	lmcache/v1/multiprocess/posix_shm.py	149	Open an existing shared-memory segment and return its address. ``nbytes`` must match the segment's actual size; ``m	STRING
HIGH	lmcache/v1/multiprocess/posix_shm.py	251	Open an existing segment as an independent ``mmap.mmap`` object. Convenience helper for non-GPU SHM transports that	STRING
214 more matches not shown…

Excessive Try-Catch Wrapping844 hits · 878 pts

Severity	File	Line	Snippet	Context
LOW	tests/conftest.py	501	except Exception as e:	CODE
LOW	tests/conftest.py	689	except Exception as e:	CODE
LOW⚡	tests/conftest.py	779	except Exception as e:	CODE
MEDIUM⚡	tests/conftest.py	781	print("Error during close obj:%s - %s", obj, e)	CODE
LOW	tests/conftest.py	917	except Exception as e:	CODE
LOW	tests/v1/test_nixl_storage.py	59	except Exception:	CODE
LOW	tests/v1/test_nixl_storage.py	259	except Exception:	CODE
MEDIUM	tests/v1/test_address_manager.py	821	def alloc_worker():	CODE
MEDIUM	tests/v1/test_address_manager.py	834	def free_worker():	CODE
MEDIUM	tests/v1/test_address_manager.py	991	def batch_alloc_worker():	CODE
MEDIUM	tests/v1/test_address_manager.py	1003	def free_worker():	CODE
LOW	tests/v1/test_address_manager.py	737	except Exception as e:	CODE
LOW	tests/v1/test_address_manager.py	784	except Exception as e:	CODE
LOW	tests/v1/test_address_manager.py	831	except Exception as e:	CODE
LOW	tests/v1/test_address_manager.py	843	except Exception as e:	CODE
LOW	tests/v1/test_address_manager.py	923	except Exception as e:	CODE
LOW	tests/v1/test_address_manager.py	960	except Exception as e:	CODE
LOW	tests/v1/test_address_manager.py	1000	except Exception as e:	CODE
LOW	tests/v1/test_address_manager.py	1012	except Exception as e:	CODE
MEDIUM	tests/v1/utils.py	197	print(f"Error during close pending tasks: - {e}")	CODE
LOW	tests/v1/utils.py	47	except Exception:	CODE
LOW	tests/v1/utils.py	70	except Exception:	CODE
LOW	tests/v1/utils.py	103	except Exception:	CODE
LOW	tests/v1/utils.py	123	except Exception:	CODE
LOW	tests/v1/utils.py	196	except Exception as e:	CODE
LOW	tests/v1/test_fused_rerope.py	22	except Exception: # allow running as a plain script (bench) without pytest	CODE
LOW	tests/v1/test_fused_rerope.py	36	except Exception: # CPU-only / no built cuda_ops (e.g. the unit-test runner)	CODE
LOW	tests/v1/test_remote_mla_worker_id_as0.py	181	except Exception as e:	CODE
MEDIUM	tests/v1/test_remote_mla_worker_id_as0.py	182	print(f"Error during shutdown: {e}")	CODE
MEDIUM	tests/v1/test_atom_mp_adapter.py	587	def submit() -> None:	CODE
LOW	tests/v1/test_atom_mp_adapter.py	743	except Exception as error:	CODE
MEDIUM	tests/v1/multiprocess/test_session.py	231	def writer() -> None:	CODE
LOW	tests/v1/multiprocess/test_session.py	200	except Exception as exc:	CODE
LOW	tests/v1/multiprocess/test_session.py	228	except Exception as exc:	CODE
LOW	tests/v1/multiprocess/test_session.py	236	except Exception as exc:	CODE
LOW	tests/v1/multiprocess/test_cache_server.py	51	except Exception:	CODE
LOW⚡	tests/v1/multiprocess/test_cache_server.py	387	except Exception as e:	CODE
MEDIUM⚡	tests/v1/multiprocess/test_cache_server.py	388	print(f"Error during unregister: {e}")	CODE
MEDIUM	tests/v1/multiprocess/test_cache_server.py	46	def _has_working_new_shared_cuda() -> bool:	CODE
LOW	tests/v1/multiprocess/test_futures.py	26	except Exception:	CODE
LOW	tests/v1/multiprocess/test_custom_types.py	159	except Exception as e:	CODE
LOW	tests/v1/multiprocess/test_custom_types.py	281	except Exception as e:	CODE
LOW	tests/v1/multiprocess/test_mq.py	160	except Exception as e:	CODE
LOW	…1/lookup_client/test_chunk_statistics_lookup_client.py	320	except Exception:	CODE
LOW	…sts/v1/distributed/test_native_connector_l2_adapter.py	72	except Exception as e:	CODE
LOW	…sts/v1/distributed/test_native_connector_l2_adapter.py	98	except Exception as e:	CODE
LOW	…s/v1/distributed/test_valkey_l2_adapter_integration.py	129	except Exception:	CODE
LOW	…s/v1/distributed/test_valkey_l2_adapter_integration.py	138	except Exception:	CODE
LOW	…v1/distributed/test_bigtable_l2_adapter_integration.py	87	except Exception:	CODE
LOW	…v1/distributed/test_bigtable_l2_adapter_integration.py	111	except Exception:	CODE
LOW	tests/v1/distributed/test_s3_l2_adapter.py	205	except Exception as e:	CODE
LOW	tests/v1/distributed/test_s3_l2_adapter.py	240	except Exception:	CODE
LOW	tests/v1/distributed/test_s3_l2_adapter.py	282	except Exception as e:	CODE
LOW	tests/v1/distributed/test_s3_l2_adapter.py	286	except Exception:	CODE
LOW	tests/v1/distributed/test_aerospike_l2_integration.py	47	except Exception:	CODE
MEDIUM	tests/v1/distributed/test_l1_manager.py	1768	def worker(thread_id):	CODE
LOW	tests/v1/distributed/test_l1_manager.py	1788	except Exception as e:	CODE
LOW	tests/v1/distributed/test_raw_block_l2_adapter.py	39	except Exception:	CODE
MEDIUM	tests/v1/distributed/test_raw_block_l2_adapter.py	33	def _has_ext() -> bool:	CODE
MEDIUM	tests/v1/distributed/test_l2_adapter_factory.py	32	def _has_raw_block_ext() -> bool:	CODE
784 more matches not shown…

Cross-File Repetition175 hits · 875 pts

Severity	File	Line	Snippet	Context
HIGH	tests/v1/test_gpu_connector.py	0	:param int size: the size of the pinned memory in bytes.	STRING
HIGH	tests/v1/test_xpu_connector.py	0	:param int size: the size of the pinned memory in bytes.	STRING
HIGH	lmcache/v1/memory_allocators/pin_memory_allocator.py	0	:param int size: the size of the pinned memory in bytes.	STRING
HIGH	lmcache/v1/memory_allocators/host_memory_allocator.py	0	:param int size: the size of the pinned memory in bytes.	STRING
HIGH	tests/v1/test_vllm_mp_adapter.py	0	restore the process-global isolated-ipc switch after the test.	STRING
HIGH	tests/v1/platform/test_cuda_ipc_wrapper.py	0	restore the process-global isolated-ipc switch after the test.	STRING
HIGH	tests/v1/platform/test_event_ipc.py	0	restore the process-global isolated-ipc switch after the test.	STRING
HIGH	tests/v1/distributed/test_mock_l2_adapter.py	0	listener that records all events for inspection in tests.	STRING
HIGH	tests/v1/distributed/test_nixl_store_l2_adapter.py	0	listener that records all events for inspection in tests.	STRING
HIGH	…s/v1/distributed/test_nixl_store_dynamic_l2_adapter.py	0	listener that records all events for inspection in tests.	STRING
HIGH	tests/v1/distributed/test_l1_manager.py	0	create a large memorylayoutdesc that will exhaust small memory. each allocation is 8mb (2m elements * 4 bytes).	STRING
HIGH	…sts/v1/distributed/test_distributed_storage_manager.py	0	create a large memorylayoutdesc that will exhaust small memory. each allocation is 8mb (2m elements * 4 bytes).	STRING
HIGH	tests/v1/distributed/test_l1_memory_manager.py	0	create a large memorylayoutdesc that will exhaust small memory. each allocation is 8mb (2m elements * 4 bytes).	STRING
HIGH	…sts/v1/mp_observability/subscribers/metrics/test_l2.py	0	capture counters before the test; yield a callable that returns deltas.	STRING
HIGH	…mp_observability/subscribers/metrics/test_cb_server.py	0	capture counters before the test; yield a callable that returns deltas.	STRING
HIGH	…rvability/subscribers/metrics/test_l1_eviction_loop.py	0	capture counters before the test; yield a callable that returns deltas.	STRING
HIGH	…v1/mp_observability/subscribers/metrics/test_lookup.py	0	capture counters before the test; yield a callable that returns deltas.	STRING
HIGH	lmcache/v1/memory_allocators/pin_memory_allocator.py	0	allocate one gpu-backed memory object. args: shapes: logical tensor shape or shapes to allocate. dtypes: logical tensor	STRING
HIGH	lmcache/v1/memory_allocators/host_memory_allocator.py	0	allocate one gpu-backed memory object. args: shapes: logical tensor shape or shapes to allocate. dtypes: logical tensor	STRING
HIGH	lmcache/v1/memory_allocators/gpu_memory_allocator.py	0	allocate one gpu-backed memory object. args: shapes: logical tensor shape or shapes to allocate. dtypes: logical tensor	STRING
HIGH	lmcache/v1/memory_allocators/pin_memory_allocator.py	0	allocate multiple gpu-backed memory objects. args: shapes: logical tensor shape or shapes for each allocation. dtypes: l	STRING
HIGH	lmcache/v1/memory_allocators/host_memory_allocator.py	0	allocate multiple gpu-backed memory objects. args: shapes: logical tensor shape or shapes for each allocation. dtypes: l	STRING
HIGH	lmcache/v1/memory_allocators/gpu_memory_allocator.py	0	allocate multiple gpu-backed memory objects. args: shapes: logical tensor shape or shapes for each allocation. dtypes: l	STRING
HIGH	lmcache/v1/memory_allocators/pin_memory_allocator.py	0	free one gpu-backed memory object. args: memory_obj: memory object to release. allocator_type: optional allocator type s	STRING
HIGH	lmcache/v1/memory_allocators/host_memory_allocator.py	0	free one gpu-backed memory object. args: memory_obj: memory object to release. allocator_type: optional allocator type s	STRING
HIGH	lmcache/v1/memory_allocators/gpu_memory_allocator.py	0	free one gpu-backed memory object. args: memory_obj: memory object to release. allocator_type: optional allocator type s	STRING
HIGH	lmcache/v1/memory_allocators/pin_memory_allocator.py	0	free multiple gpu-backed memory objects. args: memory_objs: memory objects to release. allocator_type: optional allocato	STRING
HIGH	lmcache/v1/memory_allocators/host_memory_allocator.py	0	free multiple gpu-backed memory objects. args: memory_objs: memory objects to release. allocator_type: optional allocato	STRING
HIGH	lmcache/v1/memory_allocators/gpu_memory_allocator.py	0	free multiple gpu-backed memory objects. args: memory_objs: memory objects to release. allocator_type: optional allocato	STRING
HIGH	lmcache/v1/multiprocess/engine_module.py	0	return the shared engine context. exposed for testing only.	STRING
HIGH	lmcache/v1/multiprocess/modules/p2p_controller.py	0	return the shared engine context. exposed for testing only.	STRING
HIGH	…ache/v1/multiprocess/modules/engine_driven_transfer.py	0	return the shared engine context. exposed for testing only.	STRING
HIGH	lmcache/v1/multiprocess/modules/lookup.py	0	return the shared engine context. exposed for testing only.	STRING
HIGH	…che/v1/multiprocess/modules/lmcache_driven_transfer.py	0	return the shared engine context. exposed for testing only.	STRING
HIGH	lmcache/v1/multiprocess/modules/management.py	0	return the shared engine context. exposed for testing only.	STRING
HIGH	lmcache/v1/multiprocess/modules/experimental/qstore.py	0	return the shared engine context. exposed for testing only.	STRING
HIGH	…ache/v1/multiprocess/modules/engine_driven_transfer.py	0	return handler specs for all request types this module serves. returns: a list of handlerspec entries mapping request ty	STRING
HIGH	…che/v1/multiprocess/modules/lmcache_driven_transfer.py	0	return handler specs for all request types this module serves. returns: a list of handlerspec entries mapping request ty	STRING
HIGH	lmcache/v1/multiprocess/modules/management.py	0	return handler specs for all request types this module serves. returns: a list of handlerspec entries mapping request ty	STRING
HIGH	lmcache/v1/multiprocess/modules/experimental/qstore.py	0	return handler specs for all request types this module serves. returns: a list of handlerspec entries mapping request ty	STRING
HIGH	…ache/v1/multiprocess/modules/engine_driven_transfer.py	0	refresh the worker's last-seen time and mark it ping-proven. a no-op if the instance is not tracked. args: instance_id:	STRING
HIGH	…che/v1/multiprocess/modules/lmcache_driven_transfer.py	0	refresh the worker's last-seen time and mark it ping-proven. a no-op if the instance is not tracked. args: instance_id:	STRING
HIGH	lmcache/v1/multiprocess/modules/experimental/qstore.py	0	refresh the worker's last-seen time and mark it ping-proven. a no-op if the instance is not tracked. args: instance_id:	STRING
HIGH	…ache/v1/multiprocess/modules/engine_driven_transfer.py	0	return the number of currently registered instances.	STRING
HIGH	…che/v1/multiprocess/modules/lmcache_driven_transfer.py	0	return the number of currently registered instances.	STRING
HIGH	lmcache/v1/multiprocess/modules/experimental/qstore.py	0	return the number of currently registered instances.	STRING
HIGH	…ache/v1/multiprocess/modules/engine_driven_transfer.py	0	reap q ring registrations that have gone silent. a ping-proven instance is judged against ``reap_timeout_s``; one that h	STRING
HIGH	…che/v1/multiprocess/modules/lmcache_driven_transfer.py	0	reap q ring registrations that have gone silent. a ping-proven instance is judged against ``reap_timeout_s``; one that h	STRING
HIGH	lmcache/v1/multiprocess/modules/experimental/qstore.py	0	reap q ring registrations that have gone silent. a ping-proven instance is judged against ``reap_timeout_s``; one that h	STRING
HIGH	lmcache/v1/multiprocess/protocols/controller.py	0	returns protocol definitions for debug operations. returns: dictionary mapping request names to their protocol definitio	STRING
HIGH	lmcache/v1/multiprocess/protocols/p2p.py	0	returns protocol definitions for debug operations. returns: dictionary mapping request names to their protocol definitio	STRING
HIGH	lmcache/v1/multiprocess/protocols/observability.py	0	returns protocol definitions for debug operations. returns: dictionary mapping request names to their protocol definitio	STRING
HIGH	lmcache/v1/multiprocess/protocols/engine.py	0	returns protocol definitions for debug operations. returns: dictionary mapping request names to their protocol definitio	STRING
HIGH	lmcache/v1/multiprocess/protocols/debug.py	0	returns protocol definitions for debug operations. returns: dictionary mapping request names to their protocol definitio	STRING
HIGH	…e/v1/mp_observability/subscribers/logging/mp_server.py	0	l1 logging subscriber — debug logs for l1manager events. logs are emitted via python's standard logging module. when ope	STRING
HIGH	lmcache/v1/mp_observability/subscribers/logging/sm.py	0	l1 logging subscriber — debug logs for l1manager events. logs are emitted via python's standard logging module. when ope	STRING
HIGH	lmcache/v1/mp_observability/subscribers/logging/l1.py	0	l1 logging subscriber — debug logs for l1manager events. logs are emitted via python's standard logging module. when ope	STRING
HIGH	lmcache/v1/platform/cuda/pin_memory.py	0	unpin a previously pinned host memory region. args: ptr: raw pointer (data_ptr) to the memory region. returns: true if u	STRING
HIGH	lmcache/v1/platform/base/pin_memory.py	0	unpin a previously pinned host memory region. args: ptr: raw pointer (data_ptr) to the memory region. returns: true if u	STRING
HIGH	lmcache/v1/platform/base/device_spec.py	0	unpin a previously pinned host memory region. args: ptr: raw pointer (data_ptr) to the memory region. returns: true if u	STRING
115 more matches not shown…

Unused Imports778 hits · 741 pts

Severity	File	Line	Context
LOW	tools/translate_docs_zh.py	5	CODE
LOW	tests/v1/utils.py	122	CODE
LOW	tests/v1/test_trtllm_integration.py	33	CODE
LOW	tests/v1/mp_coordinator/test_instances_usage_api.py	8	CODE
LOW	tests/v1/multiprocess/test_raw_block_l2_adapter.py	4	CODE
LOW	tests/v1/multiprocess/test_engine_driven_transfer.py	384	CODE
LOW	tests/v1/multiprocess/test_engine_driven_transfer.py	644	CODE
LOW	tests/v1/distributed/conftest.py	7	CODE
LOW	tests/v1/distributed/test_aerospike_l2_integration.py	54	CODE
LOW	tests/v1/distributed/test_memory_capacity.py	24	CODE
LOW	tests/v1/distributed/test_raw_block_l2_adapter.py	36	CODE
LOW	tests/v1/distributed/test_l2_adapter_factory.py	35	CODE
LOW	tests/v1/distributed/test_storage_manager_l2_keys.py	20	CODE
LOW	tests/v1/distributed/test_storage_manager_l2_keys.py	21	CODE
LOW	tests/v1/distributed/serde/test_asym_k16_v8_v_only.py	27	CODE
LOW	tests/v1/distributed/serde/test_asym_k16_v8_multi.py	12	CODE
LOW	tests/v1/distributed/serde/test_multi.py	25	CODE
LOW	…p_observability/subscribers/metrics/counter_helpers.py	13	CODE
LOW	tests/v1/platform/musa/test_musa_event_ordering.py	52	CODE
LOW	tests/v1/platform/rbln/test_rbln_kv_layout.py	26	CODE
LOW	tests/v1/platform/rbln/test_rbln_kv_layout.py	26	CODE
LOW	tests/v1/storage_backend/test_storage_plugin.py	28	CODE
LOW	tests/v1/storage_backend/test_raw_block_device.py	4	CODE
LOW	tests/v1/storage_backend/test_rust_raw_block_backend.py	4	CODE
LOW	tests/v1/storage_backend/test_rust_raw_block_backend.py	50	CODE
LOW	tests/v1/storage_backend/test_nixl_shared_pool.py	45	CODE
LOW	tests/v1/storage_backend/test_nixl_shared_pool.py	46	CODE
LOW	tests/v1/storage_backend/test_raw_block_core.py	4	CODE
LOW	tests/v1/storage_backend/raw_block_test_utils.py	4	CODE
LOW	tests/v1/storage_backend/test_raw_block_uring_cmd.py	58	CODE
LOW	tests/cli/commands/trace/test_dispatch.py	6	CODE
LOW	tests/cli/commands/trace/test_driver.py	13	CODE
LOW	setup_extensions/__init__.py	14	CODE
LOW	setup_extensions/__init__.py	15	CODE
LOW	setup_extensions/__init__.py	15	CODE
LOW	setup_extensions/__init__.py	16	CODE
LOW	setup_extensions/common_cpp.py	17	CODE
LOW	setup_extensions/common_cpp.py	20	CODE
LOW	setup_extensions/storage_backend_profiles/aerospike.py	16	CODE
LOW	setup_extensions/storage_backend_profiles/__init__.py	15	CODE
LOW	setup_extensions/storage_backend_profiles/mooncake.py	15	CODE
LOW	setup_extensions/build_profiles/rocm.py	16	CODE
LOW	setup_extensions/build_profiles/__init__.py	15	CODE
LOW	setup_extensions/build_profiles/__init__.py	18	CODE
LOW	setup_extensions/build_profiles/musa.py	13	CODE
LOW	setup_extensions/build_profiles/sycl.py	15	CODE
LOW	setup_extensions/build_profiles/cuda.py	16	CODE
LOW	setup_extensions/build_profiles/ascend.py	15	CODE
LOW	…amples/online_session/openai_chat_completion_client.py	14	CODE
LOW	examples/lmc_external_l2_adapter/tests/test_plugin.py	44	CODE
LOW	…rnal_l2_adapter/src/lmc_external_l2_adapter/adapter.py	19	CODE
LOW	…nal_l2_adapter/src/lmc_external_l2_adapter/__init__.py	12	CODE
LOW	…nal_l2_adapter/src/lmc_external_l2_adapter/__init__.py	12	CODE
LOW	…onnector/src/lmc_external_native_connector/__init__.py	22	CODE
LOW	…nnector/src/lmc_external_native_connector/connector.py	29	CODE
LOW	…rks/storage_backend_io/storage_backend_io_benchmark.py	15	CODE
LOW	benchmarks/multi_doc_qa/shuffle_doc_qa.py	15	CODE
LOW	.github/scripts/assert_native_ops.py	18	CODE
LOW	lmcache/usage_context.py	6	CODE
LOW	lmcache/usage_context.py	6	CODE
718 more matches not shown…

Over-Commented Block325 hits · 317 pts

Severity	File	Line	Snippet	Context
LOW	pyproject_cli.toml	61		COMMENT
LOW	pyproject_cli.toml	101	[tool.ruff.lint.isort]	COMMENT
LOW	pyproject.toml	81	# "UP",	COMMENT
LOW	pyproject.toml	121	platform = "linux"	COMMENT
LOW	pyproject.toml	161	build = "cp3-manylinux_x86_64 cp3-manylinux_aarch64"	COMMENT
LOW	csrc/engine_kv_format.h	161	#define LMC_KV_FORMAT_HD __host__ __device__	COMMENT
LOW	csrc/kv_transfer_plan_types.h	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/kv_transfer_plan_types.h	21	#endif	COMMENT
LOW	csrc/storage_backends/event_notifier.h	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/storage_backends/connector_base.h	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/storage_backends/connector_types.h	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/storage_backends/mooncake/connector.cpp	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/storage_backends/mooncake/connector.h	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/storage_backends/redis/connector.h	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/storage_backends/redis/connector.h	61	// - model_name: 25-50 chars (e.g., "meta-llama/Llama-3-70b-instruct")	COMMENT
LOW	csrc/storage_backends/aerospike/connector.cpp	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/storage_backends/fs/connector.cpp	21	size_t pos = 0;	COMMENT
LOW	csrc/storage_backends/fs/connector.h	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/cuda/mem_alloc.cpp	1	#include <cuda_runtime.h>	COMMENT
LOW	csrc/cuda/pybind.cpp	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/cuda/event_recorder.h	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/cuda/completion_recorder.h	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/cuda/event_recorder.cpp	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/lmcache_native/pybind.cpp	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/lmcache_native/periodic_event_notifier.h	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/sycl/mem_kernels_sycl.cpp	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/sycl/mem_kernels_sycl.cpp	21	// work-group handles both K and V, halving dispatch count and	COMMENT
LOW	csrc/sycl/mem_kernels_sycl.cpp	1021	offset += max_nbytes;	COMMENT
LOW	csrc/sycl/pybind_sycl.cpp	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/sycl/ac_enc_sycl.cpp	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/sycl/ac_enc_sycl.cpp	21	// (lid * BLOCK_SIZE + tx)	COMMENT
LOW	csrc/sycl/pos_kernels_sycl.cpp	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/sycl/pos_kernels_sycl.cpp	21	//	COMMENT
LOW	csrc/sycl/cal_cdf_sycl.cpp	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/sycl/cal_cdf_sycl.cpp	21	// - sub_group_size locked to 16 (native SIMD on Intel discrete GPUs).	COMMENT
LOW	csrc/sycl/ac_dec_sycl.cpp	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	csrc/sycl/ac_dec_sycl.cpp	21	//	COMMENT
LOW	csrc/sycl/cachegen_kernels_sycl.h	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	rust/raw_block/src/lib.rs	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	rust/raw_block/src/lib.rs	801	struct UringNotify {	COMMENT
LOW	rust/raw_block/src/lib.rs	821	/// eventfds into the epoll instance. Cleans up partially-built state	COMMENT
LOW	rust/raw_block/src/lib.rs	861	// We use EPOLLIN: wake when the counter becomes non-zero	COMMENT
LOW	rust/raw_block/src/lib.rs	961	}	COMMENT
LOW	rust/raw_block/src/lib.rs	1041	// Background worker thread handle	COMMENT
LOW	rust/raw_block/src/lib.rs	1481	}	COMMENT
LOW	rust/raw_block/src/lib.rs	1701	// - Check how much space is available in the ring (max 256 entries)	COMMENT
LOW	rust/raw_block/src/lib.rs	2401	/// The method waits on a per-batch condition variable that gets signaled	COMMENT
LOW	tests/v1/test_mem_kernels.py	861		COMMENT
LOW	tests/v1/test_cache_interface.py	1	# SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	tests/v1/test_vllm_integration.py	1	# SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	tests/v1/distributed/conftest.py	101	fallback_module_any.TTLLock = TTLLock	COMMENT
LOW	tests/v1/distributed/test_bitmap_ops.py	541	# might trip on. 300 chunks x 4 ranks x 5 groups.	COMMENT
LOW	tests/v1/distributed/test_valkey_l2_adapter.py	21		COMMENT
LOW	…mp_observability/subscribers/metrics/test_cb_server.py	221		COMMENT
LOW	tests/v1/lmcache_native/test_pattern_matcher.py	621	end = [9]	COMMENT
LOW	tests/v1/shm_allocator/shmfile_basic_check.yaml	1	# LMCache ShmFile Connector - basic_check configuration	COMMENT
LOW	tests/v1/shm_allocator/csrc/shm_file_worker.cpp	1	// SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	tests/v1/shm_allocator/csrc/shm_file_worker.cpp	21	#include <cstdint>	COMMENT
LOW	…1/internal_api_server/test_scripts/multiple_imports.py	1	# SPDX-License-Identifier: Apache-2.0	COMMENT
LOW	…/v1/internal_api_server/test_scripts/default_result.py	1	# SPDX-License-Identifier: Apache-2.0	COMMENT
265 more matches not shown…

Deep Nesting337 hits · 272 pts

Severity	File	Line	Context
LOW	tests/v1/test_torch_ops.py	808	CODE
LOW	tests/v1/test_torch_ops.py	1172	CODE
LOW	tests/v1/test_torch_ops.py	1418	CODE
LOW	tests/v1/test_torch_ops.py	2958	CODE
LOW	tests/v1/test_cuda_ops_parity.py	130	CODE
LOW	tests/v1/test_mp_mem_kernels.py	86	CODE
LOW	tests/v1/test_mp_mem_kernels.py	130	CODE
LOW	tests/v1/test_mp_mem_kernels.py	197	CODE
LOW	tests/v1/test_address_manager.py	811	CODE
LOW	tests/v1/test_address_manager.py	892	CODE
LOW	tests/v1/test_address_manager.py	981	CODE
LOW	tests/v1/test_address_manager.py	821	CODE
LOW	tests/v1/test_address_manager.py	834	CODE
LOW	tests/v1/test_address_manager.py	904	CODE
LOW	tests/v1/test_address_manager.py	991	CODE
LOW	tests/v1/test_address_manager.py	1003	CODE
LOW	tests/v1/utils.py	34	CODE
LOW	tests/v1/utils.py	187	CODE
LOW	tests/v1/utils.py	263	CODE
LOW	tests/v1/utils.py	419	CODE
LOW	tests/v1/utils.py	652	CODE
LOW	tests/v1/test_freeze_mode_integration.py	27	CODE
LOW	tests/v1/multiprocess/test_optimized_lookup.py	538	CODE
LOW	tests/v1/multiprocess/test_optimized_lookup.py	759	CODE
LOW	tests/v1/distributed/test_s3_l2_adapter.py	166	CODE
LOW	tests/v1/distributed/test_l1_manager.py	1758	CODE
LOW	tests/v1/distributed/test_l1_manager.py	1768	CODE
LOW	tests/v1/distributed/test_l1_l2_state_metrics.py	136	CODE
LOW	tests/v1/distributed/test_bigtable_l2_adapter.py	63	CODE
LOW	tests/v1/distributed/test_bigtable_l2_adapter.py	88	CODE
LOW	…sts/v1/mp_observability/subscribers/metrics/test_l2.py	44	CODE
LOW	…sts/v1/mp_observability/subscribers/metrics/test_l2.py	69	CODE
LOW	…v1/mp_observability/subscribers/metrics/test_engine.py	65	CODE
LOW	…/v1/mp_observability/subscribers/metrics/otel_setup.py	30	CODE
LOW	…/v1/mp_observability/subscribers/metrics/otel_setup.py	60	CODE
LOW	…observability/subscribers/metrics/test_sm_lifecycle.py	73	CODE
LOW	…v1/mp_observability/subscribers/metrics/test_lookup.py	34	CODE
LOW	…v1/mp_observability/subscribers/metrics/test_lookup.py	55	CODE
LOW	…sts/v1/mp_observability/subscribers/metrics/test_l1.py	32	CODE
LOW	…p_observability/subscribers/metrics/counter_helpers.py	31	CODE
LOW	tests/v1/cache_controller/conftest.py	218	CODE
LOW	tests/v1/cache_controller/conftest.py	248	CODE
LOW	tests/v1/cache_controller/conftest.py	323	CODE
LOW	tests/v1/cache_controller/test_registry_tree.py	451	CODE
LOW	tests/v1/cache_controller/test_registry_tree.py	519	CODE
LOW	tests/v1/cache_controller/test_registry_tree.py	493	CODE
LOW	tests/v1/cache_controller/test_registry_tree.py	541	CODE
LOW	tests/v1/lmcache_native/test_periodic_event_notifier.py	49	CODE
LOW	tests/v1/lmcache_native/test_ttl_lock.py	366	CODE
LOW	tests/v1/lmcache_native/test_ttl_lock.py	373	CODE
LOW	tests/v1/storage_backend/test_local_cpu_backend.py	339	CODE
LOW	…v1/storage_backend/test_p2p_backend_with_controller.py	134	CODE
LOW	…v1/storage_backend/test_p2p_backend_with_controller.py	231	CODE
LOW	…v1/storage_backend/test_p2p_backend_with_controller.py	440	CODE
LOW	…v1/storage_backend/test_p2p_backend_with_controller.py	561	CODE
LOW	…v1/storage_backend/test_p2p_backend_with_controller.py	608	CODE
LOW	tests/v1/storage_backend/test_dax_backend.py	142	CODE
LOW	tests/v1/storage_backend/test_dax_backend.py	1114	CODE
LOW	tests/v1/storage_backend/test_dax_backend.py	1290	CODE
LOW	tests/v1/storage_backend/test_dax_backend.py	1367	CODE
277 more matches not shown…

Self-Referential Comments71 hits · 219 pts

Severity	File	Line	Snippet	Context
MEDIUM	tests/conftest.py	284	# Create a shared store but separate instances for master/slave	COMMENT
MEDIUM	tests/conftest.py	646	# Create a fake hf3fs_fuse module if it doesn't exist	COMMENT
MEDIUM⚡	tests/v1/test_basic_check.py	48	# Create a fake check_mode_ module	COMMENT
MEDIUM⚡	tests/v1/test_memory_management.py	564	# Create a simple allocator for testing	COMMENT
MEDIUM⚡	tests/v1/test_memory_management.py	569	# Create a real TensorMemoryObj	COMMENT
MEDIUM	tests/v1/test_memory_management.py	427	# Create a memory object	COMMENT
MEDIUM	tests/v1/test_memory_management.py	473	# Create a mock memory object for testing	COMMENT
MEDIUM	tests/v1/test_memory_management.py	502	# Create a mock memory object	COMMENT
MEDIUM	tests/v1/test_config.py	809	# Create a config and update from env	COMMENT
MEDIUM	tests/v1/test_cache_engine.py	1424	# Create an engine instance	COMMENT
MEDIUM	tests/v1/test_cache_engine.py	1877	# Create a mock engine	COMMENT
MEDIUM	tests/v1/test_cache_engine.py	1969	# Create a mock engine	COMMENT
MEDIUM⚡	tests/v1/multiprocess/test_futures.py	263	# Create the raw future that will return (event_bytes, result_value)	COMMENT
MEDIUM	tests/v1/multiprocess/test_custom_types.py	31	# Create a sample IPCCacheServerKey	COMMENT
MEDIUM	tests/v1/multiprocess/test_custom_types.py	84	# Create a sample tensor	COMMENT
MEDIUM	tests/v1/multiprocess/test_custom_types.py	184	# Create a tensor with known values	COMMENT
MEDIUM	tests/v1/multiprocess/test_custom_types.py	204	# Create a queue for results	COMMENT
MEDIUM	tests/v1/multiprocess/test_mq.py	511	# Create a single test key	COMMENT
MEDIUM	tests/v1/multiprocess/test_mq.py	535	# Create a different test key	COMMENT
MEDIUM⚡	tests/v1/distributed/test_mock_l2_adapter.py	625	# Create a larger object to see noticeable delay	COMMENT
MEDIUM	tests/v1/cache_controller/conftest.py	225	# Create a mock registry with all required methods	COMMENT
MEDIUM	tests/v1/cache_controller/test_config.py	55	# Create a temporary YAML config file	COMMENT
MEDIUM	tests/v1/plugin/test_runtime_plugin_launcher.py	96	# Create a Python plugin file	COMMENT
MEDIUM	tests/v1/plugin/test_runtime_plugin_launcher.py	238	# Create a Python file with shebang	COMMENT
MEDIUM	tests/v1/plugin/test_runtime_plugin_launcher.py	258	# Create a Bash file	COMMENT
MEDIUM	tests/v1/plugin/test_runtime_plugin_launcher.py	278	# Create a Python file without shebang	COMMENT
MEDIUM	tests/v1/plugin/test_runtime_plugin_launcher.py	301	# Create an unsupported file type	COMMENT
MEDIUM	tests/v1/lmcache_native/test_ttl_lock.py	115	# Create a lock with 1 second TTL	COMMENT
MEDIUM	…ts/v1/storage_backend/test_pd_backend_bidirectional.py	303	# Create a mock channel	COMMENT
MEDIUM	…ts/v1/storage_backend/test_pd_backend_bidirectional.py	418	# Create a query for 3 keys (2 cached, 1 not)	COMMENT
MEDIUM	tests/v1/storage_backend/test_storage_plugin.py	300	# Create a test key	COMMENT
MEDIUM⚡	tests/v1/storage_backend/test_gds_backend.py	238	# Create a path that doesn't exist	COMMENT
MEDIUM	tests/v1/storage_backend/test_gds_backend.py	289	# Create a dummy file so os.path.exists returns True	COMMENT
MEDIUM	tests/v1/storage_backend/test_gds_backend.py	315	# Create a dummy file	COMMENT
MEDIUM	tests/v1/storage_backend/test_gds_backend.py	341	# Create a dummy file	COMMENT
MEDIUM	tests/v1/storage_backend/test_hf3fs_connector.py	180	# Create the 3fs-virt directory structure that HF3FS expects	COMMENT
MEDIUM	tests/v1/storage_backend/test_storage_manager.py	189	# Create a mock future that returns the result	COMMENT
MEDIUM	tests/v1/storage_backend/test_storage_manager.py	233	# Create a mock future that returns the result	COMMENT
MEDIUM	tests/v1/storage_backend/test_storage_manager.py	285	# Create a mock future that returns the result	COMMENT
MEDIUM	tests/v1/storage_backend/test_storage_manager.py	332	# Create a mock future that returns the result	COMMENT
MEDIUM	tests/v1/storage_backend/test_storage_manager.py	369	# Create a mock future that returns the result	COMMENT
MEDIUM	tests/v1/storage_backend/test_bigtable_integration.py	269	# Create a payload of ~8.38 MB (exceeds the 5.0 MB threshold)	COMMENT
MEDIUM	tests/v1/internal_api_server/test_cache_clear.py	23	# Create a mock LMCache engine	COMMENT
MEDIUM	tests/disagg/test_nixl_storage_backend.py	81	"nixl_path": tempfile.mkdtemp(), # Create a temporary directory for testing	CODE
MEDIUM	tests/disagg/test_nixl_cache_engine.py	191	# Create the VLLMPagedMemGPUConnectorV2	COMMENT
MEDIUM	tests/disagg/test_nixl_cache_engine.py	202	# Create the LMCacheEngine (will be reused across rounds)	COMMENT
MEDIUM	tests/disagg/test_nixl_channel.py	128	# Create the NixlChannel	COMMENT
MEDIUM	examples/blend_in_process/blend.py	146	# Define the shared prompt and specific prompts	COMMENT
MEDIUM	…/kv_cache_reuse/local_backends/rust_backend_offload.py	158	# Define the shared prompt and specific prompts	COMMENT
MEDIUM	examples/kv_cache_reuse/local_backends/offload.py	131	# Define the shared prompt and specific prompts	COMMENT
MEDIUM	…rks/storage_backend_io/storage_backend_io_benchmark.py	411	# Create the specific backend	COMMENT
MEDIUM	benchmarks/rag/rag.py	308	# Create a dataframe	COMMENT
MEDIUM	benchmarks/multi_doc_qa/multi_doc_qa.py	252	# Create the OpenAI client	COMMENT
MEDIUM	benchmarks/long_doc_qa/long_doc_qa.py	484	# Create the OpenAI client	COMMENT
MEDIUM	.github/scripts/download_gh_release_model.sh	142	# Create the refs/main pointer so snapshot_download(local_files_only=True)	COMMENT
MEDIUM	.buildkite/k3_harness/values.yaml	3	# This file is for documentation only.	COMMENT
MEDIUM	.buildkite/scripts/vllm-integration-tests.sh	782	# Create the container image	COMMENT
MEDIUM	lmcache/connections.py	2	# This file is copied from the vLLM project (https://github.com/vllm-project/vllm).	COMMENT
MEDIUM	lmcache/v1/memory_allocators/lazy_memory_allocator.py	147	# Create the tensor memory allocator	COMMENT
MEDIUM	lmcache/v1/health_monitor/base.py	167	# Create a health monitor with manager	STRING
11 more matches not shown…

Structural Annotation Overuse106 hits · 177 pts

Severity	File	Line	Snippet	Context
LOW	csrc/storage_backends/README.md	83	### Step 1: C++ connector — inherit from ConnectorBase	COMMENT
LOW	csrc/storage_backends/README.md	153	### Step 2: Pybind module	COMMENT
LOW	csrc/storage_backends/README.md	177	### Step 3: Build system — register in setup.py	COMMENT
LOW	csrc/storage_backends/README.md	199	### Step 4: Python client — non-MP mode integration	COMMENT
LOW	csrc/storage_backends/README.md	220	### Step 5: MP mode integration — L2 adapter config + factory	COMMENT
LOW⚡	tests/v1/test_health_monitor_fallback_recovery.py	386	# Step 1: Both backends healthy	COMMENT
LOW⚡	tests/v1/test_health_monitor_fallback_recovery.py	394	# Step 2: Backend A fails	COMMENT
LOW⚡	tests/v1/test_health_monitor_fallback_recovery.py	401	# Step 3: Backend B also fails	COMMENT
LOW	tests/v1/test_health_monitor_fallback_recovery.py	412	# Step 4: Backend A recovers	COMMENT
LOW	tests/v1/test_health_monitor_fallback_recovery.py	424	# Step 5: Backend B recovers	COMMENT
LOW⚡	tests/v1/test_freeze_mode_integration.py	115	# Step 1: Verify initial freeze mode is OFF	COMMENT
LOW⚡	tests/v1/test_freeze_mode_integration.py	122	# Step 2: Normal mode - Store should work	COMMENT
LOW⚡	tests/v1/test_freeze_mode_integration.py	155	# Step 3: Enable freeze mode via HTTP API	COMMENT
LOW⚡	tests/v1/test_freeze_mode_integration.py	169	# Step 4: Read-only mode - Store should be skipped	COMMENT
LOW⚡	tests/v1/test_freeze_mode_integration.py	196	# Step 5: Disable freeze mode via HTTP API	COMMENT
LOW⚡	tests/v1/test_freeze_mode_integration.py	205	# Step 6: Store should work again after disabling	COMMENT
LOW	tests/v1/test_cache_engine.py	1212	# NOTE: The default backend cache size is 2 GB.	COMMENT
LOW⚡	tests/v1/distributed/test_mock_l2_adapter.py	519	# Step 1: Store	COMMENT
LOW⚡	tests/v1/distributed/test_mock_l2_adapter.py	525	# Step 2: Lookup and lock	COMMENT
LOW⚡	tests/v1/distributed/test_mock_l2_adapter.py	531	# Step 3: Load	COMMENT
LOW⚡	tests/v1/distributed/test_mock_l2_adapter.py	540	# Step 4: Unlock	COMMENT
LOW⚡	tests/v1/distributed/test_nixl_store_l2_adapter.py	607	# Step 1: Store from page 0 (filled with 123.0)	COMMENT
LOW⚡	tests/v1/distributed/test_nixl_store_l2_adapter.py	614	# Step 2: Lookup and lock	COMMENT
LOW⚡	tests/v1/distributed/test_nixl_store_l2_adapter.py	620	# Step 3: Load into page 1 (initially 0.0)	COMMENT
LOW⚡	tests/v1/distributed/test_nixl_store_l2_adapter.py	630	# Step 4: Unlock	COMMENT
LOW	tests/v1/distributed/test_object_key_parallel.py	678	# Step 3: Scheduler lookup with interleaved keys	COMMENT
LOW⚡	tests/v1/distributed/test_object_key_parallel.py	695	# Step 4: Verify hit count	COMMENT
LOW⚡	tests/v1/distributed/test_object_key_parallel.py	703	# Step 5: Workers retrieve their chunks	COMMENT
LOW⚡	tests/v1/shm_allocator/shmfile_basic_check.yaml	4	# Step 1: Build shm_file_worker (PIPE mode, same machine)	COMMENT
LOW⚡	tests/v1/shm_allocator/shmfile_basic_check.yaml	8	# Step 2: Run basic_check	COMMENT
LOW⚡	tests/v1/internal_api_server/test_lookup_api.py	288	# Step 1: Worker side - recreate server	COMMENT
LOW⚡	tests/v1/internal_api_server/test_lookup_api.py	298	# Step 2: Scheduler side - recreate client	COMMENT
LOW⚡	tests/v1/internal_api_server/test_lookup_api.py	314	# Step 1: Dryrun - check what would be created	COMMENT
LOW⚡	tests/v1/internal_api_server/test_lookup_api.py	324	# Step 2: Close current client	COMMENT
LOW⚡	tests/v1/internal_api_server/test_lookup_api.py	329	# Step 3: Create new client	COMMENT
LOW⚡	tests/v1/internal_api_server/test_backend_api.py	232	# Step 1: close RemoteBackend	COMMENT
LOW⚡	tests/v1/internal_api_server/test_backend_api.py	237	# Step 2: update config via /conf (uses conf_api)	COMMENT
LOW⚡	tests/v1/internal_api_server/test_backend_api.py	242	# Step 3: create new backends	COMMENT
LOW⚡	tests/v1/internal_api_server/test_backend_api.py	252	# Step 4: list backends	COMMENT
LOW	docs/design/v1/multiprocess/http_api_extension.md	90	### Step 1: Create the Module	COMMENT
LOW	docs/design/v1/multiprocess/http_api_extension.md	116	### Step 2: Done	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	51	# Step 0: Generate the master key (never checked in / logged)	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	104	# Step 1: Launch lmcache MP server (CPU L1 + disk L2 with encrypt serde)	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	134	# Step 2: Launch vLLM with LMCacheMPConnector	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	178	# Step 3: First inference — cold path (L1 -> L2 store with AES-GCM encrypt)	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	224	# Step 4: Clear L1 (CPU) cache	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	234	# Step 5: Re-run the same request — triggers L2 prefetch + AES-GCM decrypt	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	254	# Step 6: Show metrics / status	COMMENT
LOW⚡	examples/serde/fp8/run_serde_fp8_example.sh	85	# Step 1: Launch lmcache MP server (CPU L1 + disk L2 with fp8 serde)	COMMENT
LOW⚡	examples/serde/fp8/run_serde_fp8_example.sh	115	# Step 2: Launch vLLM with LMCacheMPConnector	COMMENT
LOW⚡	examples/serde/fp8/run_serde_fp8_example.sh	159	# Step 3: First inference — cold path (L1 -> L2 store with fp8 serialize)	COMMENT
LOW⚡	examples/serde/fp8/run_serde_fp8_example.sh	197	# Step 4: Clear L1 (CPU) cache	COMMENT
LOW⚡	examples/serde/fp8/run_serde_fp8_example.sh	207	# Step 5: Re-run the same request — triggers L2 prefetch + fp8 deserialize	COMMENT
LOW⚡	examples/serde/fp8/run_serde_fp8_example.sh	226	# Step 6: Show metrics / status	COMMENT
LOW	examples/kv_cache_reuse/local_backends/README.md	12	# WARNING: This will erase the content of target device.	COMMENT
LOW⚡	examples/kv_cache_reuse/remote_backends/s3/README.md	7	## Step 1: Configure your S3 bucket and (optional) EC2 compute instance	COMMENT
LOW⚡	examples/kv_cache_reuse/remote_backends/s3/README.md	14	## Step 2: Fill out `example.yaml`	COMMENT
LOW⚡	examples/kv_cache_reuse/remote_backends/s3/README.md	18	## Step 3: Start an vLLM engine with LMCache	COMMENT
LOW⚡	examples/kv_cache_reuse/remote_backends/s3/README.md	24	## Step 4: Sending requests	COMMENT
LOW	…amples/kv_cache_reuse/remote_backends/valkey/README.md	11	## Step 1: Start a local Valkey	COMMENT
46 more matches not shown…

Verbosity Indicators81 hits · 138 pts

Severity	File	Line	Snippet	Context
LOW⚡	tests/v1/test_health_monitor_fallback_recovery.py	386	# Step 1: Both backends healthy	COMMENT
LOW⚡	tests/v1/test_health_monitor_fallback_recovery.py	394	# Step 2: Backend A fails	COMMENT
LOW⚡	tests/v1/test_health_monitor_fallback_recovery.py	401	# Step 3: Backend B also fails	COMMENT
LOW	tests/v1/test_health_monitor_fallback_recovery.py	412	# Step 4: Backend A recovers	COMMENT
LOW	tests/v1/test_health_monitor_fallback_recovery.py	424	# Step 5: Backend B recovers	COMMENT
LOW⚡	tests/v1/test_freeze_mode_integration.py	115	# Step 1: Verify initial freeze mode is OFF	COMMENT
LOW⚡	tests/v1/test_freeze_mode_integration.py	122	# Step 2: Normal mode - Store should work	COMMENT
LOW⚡	tests/v1/test_freeze_mode_integration.py	155	# Step 3: Enable freeze mode via HTTP API	COMMENT
LOW⚡	tests/v1/test_freeze_mode_integration.py	169	# Step 4: Read-only mode - Store should be skipped	COMMENT
LOW⚡	tests/v1/test_freeze_mode_integration.py	196	# Step 5: Disable freeze mode via HTTP API	COMMENT
LOW⚡	tests/v1/test_freeze_mode_integration.py	205	# Step 6: Store should work again after disabling	COMMENT
LOW⚡	tests/v1/distributed/test_mock_l2_adapter.py	519	# Step 1: Store	COMMENT
LOW⚡	tests/v1/distributed/test_mock_l2_adapter.py	525	# Step 2: Lookup and lock	COMMENT
LOW⚡	tests/v1/distributed/test_mock_l2_adapter.py	531	# Step 3: Load	COMMENT
LOW⚡	tests/v1/distributed/test_mock_l2_adapter.py	540	# Step 4: Unlock	COMMENT
LOW⚡	tests/v1/distributed/test_nixl_store_l2_adapter.py	607	# Step 1: Store from page 0 (filled with 123.0)	COMMENT
LOW⚡	tests/v1/distributed/test_nixl_store_l2_adapter.py	614	# Step 2: Lookup and lock	COMMENT
LOW⚡	tests/v1/distributed/test_nixl_store_l2_adapter.py	620	# Step 3: Load into page 1 (initially 0.0)	COMMENT
LOW⚡	tests/v1/distributed/test_nixl_store_l2_adapter.py	630	# Step 4: Unlock	COMMENT
LOW	tests/v1/distributed/test_object_key_parallel.py	678	# Step 3: Scheduler lookup with interleaved keys	COMMENT
LOW⚡	tests/v1/distributed/test_object_key_parallel.py	695	# Step 4: Verify hit count	COMMENT
LOW⚡	tests/v1/distributed/test_object_key_parallel.py	703	# Step 5: Workers retrieve their chunks	COMMENT
LOW⚡	tests/v1/shm_allocator/shmfile_basic_check.yaml	4	# Step 1: Build shm_file_worker (PIPE mode, same machine)	COMMENT
LOW⚡	tests/v1/shm_allocator/shmfile_basic_check.yaml	8	# Step 2: Run basic_check	COMMENT
LOW⚡	tests/v1/internal_api_server/test_lookup_api.py	288	# Step 1: Worker side - recreate server	COMMENT
LOW⚡	tests/v1/internal_api_server/test_lookup_api.py	298	# Step 2: Scheduler side - recreate client	COMMENT
LOW⚡	tests/v1/internal_api_server/test_lookup_api.py	314	# Step 1: Dryrun - check what would be created	COMMENT
LOW⚡	tests/v1/internal_api_server/test_lookup_api.py	324	# Step 2: Close current client	COMMENT
LOW⚡	tests/v1/internal_api_server/test_lookup_api.py	329	# Step 3: Create new client	COMMENT
LOW⚡	tests/v1/internal_api_server/test_backend_api.py	232	# Step 1: close RemoteBackend	COMMENT
LOW⚡	tests/v1/internal_api_server/test_backend_api.py	237	# Step 2: update config via /conf (uses conf_api)	COMMENT
LOW⚡	tests/v1/internal_api_server/test_backend_api.py	242	# Step 3: create new backends	COMMENT
LOW⚡	tests/v1/internal_api_server/test_backend_api.py	252	# Step 4: list backends	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	51	# Step 0: Generate the master key (never checked in / logged)	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	104	# Step 1: Launch lmcache MP server (CPU L1 + disk L2 with encrypt serde)	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	134	# Step 2: Launch vLLM with LMCacheMPConnector	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	178	# Step 3: First inference — cold path (L1 -> L2 store with AES-GCM encrypt)	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	224	# Step 4: Clear L1 (CPU) cache	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	234	# Step 5: Re-run the same request — triggers L2 prefetch + AES-GCM decrypt	COMMENT
LOW⚡	examples/serde/aesgcm/run_serde_aesgcm_example.sh	254	# Step 6: Show metrics / status	COMMENT
LOW⚡	examples/serde/fp8/run_serde_fp8_example.sh	85	# Step 1: Launch lmcache MP server (CPU L1 + disk L2 with fp8 serde)	COMMENT
LOW⚡	examples/serde/fp8/run_serde_fp8_example.sh	115	# Step 2: Launch vLLM with LMCacheMPConnector	COMMENT
LOW⚡	examples/serde/fp8/run_serde_fp8_example.sh	159	# Step 3: First inference — cold path (L1 -> L2 store with fp8 serialize)	COMMENT
LOW⚡	examples/serde/fp8/run_serde_fp8_example.sh	197	# Step 4: Clear L1 (CPU) cache	COMMENT
LOW⚡	examples/serde/fp8/run_serde_fp8_example.sh	207	# Step 5: Re-run the same request — triggers L2 prefetch + fp8 deserialize	COMMENT
LOW⚡	examples/serde/fp8/run_serde_fp8_example.sh	226	# Step 6: Show metrics / status	COMMENT
LOW	.buildkite/scripts/multiprocessing-test/run-mp-test.sh	46	# Step 1: Build docker images	COMMENT
LOW	.buildkite/scripts/multiprocessing-test/run-mp-test.sh	57	# Step 2: Launch containers	COMMENT
LOW	.buildkite/scripts/multiprocessing-test/run-mp-test.sh	68	# Step 3: Wait for vLLM to be ready	COMMENT
LOW	.buildkite/scripts/multiprocessing-test/run-mp-test.sh	79	# Step 4: Run lm_eval workload test	COMMENT
LOW	.buildkite/scripts/multiprocessing-test/run-mp-test.sh	90	# Step 5: Run vllm bench serve test	COMMENT
LOW	.buildkite/scripts/multiprocessing-test/run-mp-test.sh	101	# Step 6: Run long doc QA test	COMMENT
LOW	.buildkite/scripts/multiprocessing-test/run-mp-test.sh	112	# Step 7: Run fault tolerance test (destructive — kills LMCache container)	COMMENT
LOW	.buildkite/scripts/multiprocessing-test/run-mp-test.sh	127	# Step 8: Cleanup runs automatically via trap	COMMENT
LOW	…te/scripts/multiprocessing-test/run-fault-tolerance.sh	79	# Step 1: Warmup bench (measure timing for kill calibration)	COMMENT
LOW	…te/scripts/multiprocessing-test/run-fault-tolerance.sh	93	# Step 2: Bench with mid-flight LMCache kill	COMMENT
LOW	…te/scripts/multiprocessing-test/run-fault-tolerance.sh	118	# Step 3: Quick curl health check	COMMENT
LOW⚡	…te/k3_tests/multiprocess/scripts/run-long-doc-qa-l2.sh	72	# Step 1: Kill existing LMCache + vLLM, relaunch both with L2 config	COMMENT
LOW⚡	…te/k3_tests/multiprocess/scripts/run-long-doc-qa-l2.sh	228	# Step 2: Run benchmarks	COMMENT
LOW⚡	…te/k3_tests/multiprocess/scripts/run-long-doc-qa-l2.sh	256	# Step 3: Verify thresholds	COMMENT
21 more matches not shown…

Redundant / Tautological Comments75 hits · 112 pts

Severity	File	Line	Snippet	Context
LOW	docker/example_build.sh	7	# Set VLLM_VERSION to a specific version before running this script,	COMMENT
LOW⚡	tests/conftest.py	776	# Check if object has a close method	COMMENT
LOW	tests/v1/test_impl_completeness.py	73	# Check if the method is actually implemented in the class	COMMENT
LOW	tests/v1/utils.py	667	# Check if it's callable (function, method, etc.)	COMMENT
LOW	tests/v1/multiprocess/test_custom_types.py	217	# Check if process completed successfully	COMMENT
LOW	tests/v1/multiprocess/test_mq.py	272	# Check if client completed successfully	COMMENT
LOW	…v1/storage_backend/test_p2p_backend_with_controller.py	65	# Check if backend is properly initialized by checking its attributes	COMMENT
LOW	…1/storage_backend/test_local_cpu_backend_nixl_paged.py	241	# Set cpu_size_bytes to exactly 1 byte — smaller than one chunk.	COMMENT
LOW	tests/v1/internal_api_server/test_run_script.py	41	# Check if we've closed all braces	COMMENT
LOW	tests/disagg/test_nixl_cache_engine.py	82	# Check if the mean is close to the pattern value	COMMENT
LOW	tests/disagg/test_nixl_cache_engine.py	293	# Check if all tokens were retrieved	COMMENT
LOW	examples/chunk_statistics/analyze_chunk_hashes.py	393	# Print results	COMMENT
LOW	examples/kubernetes/health_probe.py	65	# Check if server responded with success	COMMENT
LOW	…rks/storage_backend_io/storage_backend_io_benchmark.py	474	# Check if this is a two-phase benchmark (write + read)	COMMENT
LOW	…rks/storage_backend_io/storage_backend_io_benchmark.py	1225	# Print results	COMMENT
LOW	benchmarks/multi_doc_qa/multi_doc_qa.py	322	# Print results	COMMENT
LOW	benchmarks/multi_round_qa/multi-round-qa.py	455	# Check if can join new user session	COMMENT
LOW	benchmarks/long_doc_qa/long_doc_qa.py	561	# Print results	COMMENT
LOW	benchmarks/microbenchmark/ttl_lock_benchmark.py	522	# Check if we can sustain 10K+ ops/sec with 4-8 threads	COMMENT
LOW	.buildkite/k3_harness/gpu-monitor.sh	67	# Check if this PID is inside any cgroup containing "kubepods"	COMMENT
LOW	.buildkite/k3_harness/setup-cluster.sh	66	# Check if image already exists in K3s containerd	COMMENT
LOW	.buildkite/scripts/vllm-integration-tests.sh	79	# Check if port is available using netstat	COMMENT
LOW	.buildkite/scripts/vllm-integration-tests.sh	669	# Check if baseline exists, skip comparisons if not	COMMENT
LOW	…ildkite/scripts/multiprocessing-test/run-vllm-bench.sh	105	# Check if result files exist	COMMENT
LOW	.buildkite/scripts/multiprocessing-test/common.sh	28	# Check if uv is available	COMMENT
LOW	…uildkite/scripts/multiprocessing-test/wait-for-vllm.sh	37	# Check if container is still running	COMMENT
LOW	…uildkite/scripts/multiprocessing-test/wait-for-vllm.sh	46	# Check if vLLM health endpoint is responding	COMMENT
LOW	…ldkite/scripts/multiprocessing-test/run-long-doc-qa.sh	118	# Check if result files exist	COMMENT
LOW	lmcache/utils.py	263	# Check if it's a range format [start,end]	COMMENT
LOW	lmcache/v1/memory_management.py	1596	# Check if free blocks are properly coalesced	COMMENT
LOW	lmcache/v1/memory_management.py	1603	# Check if total size matches	COMMENT
LOW	lmcache/v1/cache_engine.py	461	# Check if freeze mode is enabled	COMMENT
LOW	lmcache/v1/cache_engine.py	647	# Check if freeze mode is enabled	COMMENT
LOW	lmcache/v1/config_base.py	583	# Check if this configuration class has definitions with converters	COMMENT
LOW	lmcache/v1/pin_monitor.py	134	# Check if object is still pinned and has exceeded timeout	COMMENT
LOW	lmcache/v1/memory_allocators/tensor_memory_allocator.py	308	# Check if the numbers are consistent	COMMENT
LOW	lmcache/v1/memory_allocators/tensor_memory_allocator.py	317	# Check if the blocks are coalesced	COMMENT
LOW	…/v1/memory_allocators/paged_tensor_memory_allocator.py	321	# Check if the numbers are consistent	COMMENT
LOW	lmcache/v1/lookup_client/lmcache_async_lookup_client.py	165	# Check if any aborted lookups are finished, send cleanup messages	COMMENT
LOW	lmcache/v1/lookup_client/factory.py	60	# Check if external_lookup_client is configured	COMMENT
LOW	lmcache/v1/lookup_client/factory.py	78	# Check if bypass lookup is enabled and lmcache_engine is provided	COMMENT
LOW	lmcache/v1/distributed/l2_adapters/factory.py	127	# Check if the name is now registered	COMMENT
LOW	…1/mp_observability/subscribers/metrics/l1_lifecycle.py	119	# Check if this is a reuse of an evicted chunk.	COMMENT
LOW	lmcache/v1/cache_controller/config.py	128	# Check if this field has an env_converter in the definitions	COMMENT
LOW⚡	lmcache/v1/cache_controller/controller_manager.py	333	# Check if POLLIN flag is set (indicates readable/pending messages)	COMMENT
LOW	lmcache/v1/cache_controller/commands/full_sync.py	43	# Check if full sync is already in progress	COMMENT
LOW	…ache_controller/controllers/registration_controller.py	231	# Check if full sync is needed (e.g., controller restart)	COMMENT
LOW	…e/v1/cache_controller/controllers/full_sync_tracker.py	146	# Check if already syncing with different sync_id	COMMENT
LOW	…cache/v1/cache_controller/controllers/kv_controller.py	154	# Check if worker is currently in full sync	COMMENT
LOW	lmcache/v1/health_monitor/base.py	498	# Check if we have any health checks that need to run	COMMENT
LOW	lmcache/v1/storage_backend/remote_backend.py	349	# Check if local_cpu_backend is available (required for memory allocation)	COMMENT
LOW	lmcache/v1/storage_backend/remote_backend.py	401	# Check if local_cpu_backend is available (required for memory allocation)	COMMENT
LOW	lmcache/v1/storage_backend/remote_backend.py	555	# Check if local_cpu_backend is available (required for memory allocation)	COMMENT
LOW	lmcache/v1/storage_backend/gds_backend.py	1103	# Check if file is large enough for the requested read	COMMENT
LOW	lmcache/v1/storage_backend/batched_message_sender.py	109	# Check if we have messages to process while holding the lock	COMMENT
LOW	lmcache/v1/storage_backend/local_cpu_backend.py	461	# Check if io_uring is enabled for fixed buffer support	COMMENT
LOW	lmcache/v1/storage_backend/local_cpu_backend.py	475	# Check if lazy memory allocator should be enabled	COMMENT
LOW	lmcache/v1/storage_backend/nixl_storage_backend.py	1918	# Check if already in progress	COMMENT
LOW	lmcache/v1/storage_backend/raw_block/core.py	1129	# Check if the buffer pointer is aligned	COMMENT
LOW	lmcache/v1/storage_backend/connector/audit_connector.py	46	# Check if method is marked with @NotAudit	COMMENT
15 more matches not shown…

Modern Structural Boilerplate103 hits · 105 pts

Severity	File	Line	Snippet	Context
LOW	tests/v1/test_health_monitor_fallback_recovery.py	41	def set_ping_error_code(self, error_code: int) -> None:	CODE
LOW⚡	tests/v1/test_musa_support.py	79	def set_device(self, _idx: int) -> None:	CODE
LOW	tests/v1/test_health_monitor.py	45	def set_healthy(self, healthy: bool) -> None:	CODE
LOW	tests/v1/distributed/serde/test_async_processor.py	176	def set_used_size(self, n: int) -> None:	CODE
LOW	…nal_l2_adapter/src/lmc_external_l2_adapter/__init__.py	17	__all__ = ["InMemoryL2Adapter", "InMemoryL2AdapterConfig"]	CODE
LOW	…onnector/src/lmc_external_native_connector/__init__.py	26	__all__ = [	CODE
LOW	examples/remote_config_server/config_server.py	28	logger = logging.getLogger(__name__)	CODE
LOW	…rks/storage_backend_io/storage_backend_io_benchmark.py	359	def _setup_device(self) -> None:	CODE
LOW	…rks/storage_backend_io/storage_backend_io_benchmark.py	755	def _setup_device(self) -> None:	CODE
LOW	.buildkite/correctness/async_request.py	16	logger = logging.getLogger(__name__)	CODE
LOW	lmcache/usage_context.py	16	__all__ = [	CODE
LOW	lmcache/__init__.py	20	__all__ = ["__version__", "torch_dev", "torch_device_type"]	CODE
LOW	lmcache/v1/memory_management.py	297	def set_used_size(self, n: int) -> None: # noqa: B027	CODE
LOW	lmcache/v1/memory_management.py	700	def set_used_size(self, n: int) -> None:	CODE
LOW	lmcache/v1/cache_engine.py	248	def set_health_monitor(self, health_monitor: "HealthMonitor") -> None:	CODE
LOW	lmcache/v1/cache_engine.py	362	def set_hot_cache(self, enabled: bool) -> None:	CODE
LOW	lmcache/v1/mp_coordinator/controllers/__init__.py	53	__all__ = ["Controller", "build_controllers"]	CODE
LOW	lmcache/v1/mp_coordinator/views/__init__.py	48	__all__ = ["View", "base", "build_views"]	CODE
LOW	lmcache/v1/multiprocess/session.py	54	def set_tokens(self, full_token_ids: list[int]) -> None:	CODE
LOW	lmcache/v1/multiprocess/futures.py	62	def set_result(self, result: T) -> None:	CODE
LOW	lmcache/v1/multiprocess/futures.py	235	def set_result(self, result: T) -> None:	CODE
LOW	lmcache/v1/multiprocess/posix_shm.py	45	logger = logging.getLogger(__name__)	CODE
LOW	lmcache/v1/multiprocess/transfer_context/__init__.py	29	__all__ = [	CODE
LOW	lmcache/v1/multiprocess/modules/p2p_controller.py	580	def _setup_metrics(self) -> None:	CODE
LOW	lmcache/v1/multiprocess/modules/lookup.py	622	def _setup_metrics(self) -> None:	CODE
LOW⚡	lmcache/v1/multiprocess/protocols/__init__.py	121	__all__ = [	CODE
LOW	lmcache/v1/offload_server/__init__.py	11	__all__ = [	CODE
LOW	…che/v1/lookup_client/chunk_statistics_lookup_client.py	197	def _setup_metrics(self) -> None:	CODE
LOW	lmcache/v1/lookup_client/__init__.py	17	__all__ = [	CODE
LOW	lmcache/v1/lookup_client/record_strategies/__init__.py	56	__all__ = [	CODE
LOW	lmcache/v1/distributed/quota_manager.py	55	def set_quota(self, cache_salt: str, limit_bytes: int) -> None:	CODE
LOW	lmcache/v1/distributed/quota_manager.py	91	def set_default_limit_bytes(self, limit_bytes: int \| None) -> None:	CODE
LOW	lmcache/v1/distributed/memory_manager/__init__.py	24	__all__ = [	CODE
LOW	lmcache/v1/distributed/eviction_policy/__init__.py	20	__all__ = [	CODE
LOW	lmcache/v1/distributed/storage_controllers/__init__.py	36	__all__ = [	CODE
LOW	lmcache/v1/distributed/bitmap_ops/__init__.py	21	__all__ = [	CODE
LOW	lmcache/v1/distributed/transfer_channel/__init__.py	30	__all__ = [	CODE
LOW⚡	…ache/v1/distributed/l2_adapters/hfbucket_l2_adapter.py	874	def _set_cached_object_size_locked(self, object_path: str, size: int) -> None:	CODE
LOW	lmcache/v1/distributed/l2_adapters/__init__.py	64	__all__ = [	CODE
LOW	lmcache/v1/distributed/l2_adapters/base.py	363	def set_backend_identity(self, name: str, shared: bool = False) -> None:	CODE
LOW⚡	lmcache/v1/distributed/l2_adapters/serde_wrapper.py	333	def set_backend_identity(self, name: str, shared: bool = False) -> None:	CODE
LOW	…/v1/distributed/l2_adapters/fault_inject_l2_adapter.py	387	def set_backend_identity(self, name: str, shared: bool = False) -> None:	CODE
LOW	lmcache/v1/distributed/serde/__init__.py	47	__all__ = [	CODE
LOW	lmcache/v1/distributed/serde/turboquant/__init__.py	11	__all__ = [	CODE
LOW	lmcache/v1/mp_observability/trace/decorator.py	49	def set_tracing_enabled(enabled: bool) -> None:	CODE
LOW	lmcache/v1/mp_observability/trace/__init__.py	39	__all__ = [	CODE
LOW	lmcache/v1/mp_observability/subscribers/__init__.py	18	__all__ = [	CODE
LOW	…he/v1/mp_observability/subscribers/metrics/__init__.py	49	__all__ = [	CODE
LOW	…he/v1/mp_observability/subscribers/tracing/__init__.py	18	__all__ = [	CODE
LOW	…he/v1/mp_observability/subscribers/logging/__init__.py	22	__all__ = [	CODE
LOW	lmcache/v1/cache_controller/__init__.py	6	__all__ = [	CODE
LOW	lmcache/v1/cache_controller/utils.py	187	def update_seq_num(self, location: str, seq_num: int) -> None:	CODE
LOW	lmcache/v1/cache_controller/commands/__init__.py	12	__all__ = [	CODE
LOW	lmcache/v1/cache_controller/controllers/__init__.py	8	__all__ = [	CODE
LOW	…e/v1/cache_controller/controllers/full_sync_tracker.py	71	def set_need_full_sync_all(self, need: bool) -> None:	CODE
LOW	…cache/v1/cache_controller/controllers/kv_controller.py	74	def _setup_metrics(self) -> None:	CODE
LOW	lmcache/v1/platform/isolated_ipc.py	18	def set_isolated_ipc(enabled: bool) -> None:	CODE
LOW	lmcache/v1/platform/__init__.py	25	__all__ = [	CODE
LOW⚡	lmcache/v1/platform/cpu/shm.py	43	__all__ = [	CODE
LOW	lmcache/v1/platform/cpu/stub_cpu_device.py	282	def set_stream(self, stream: StubStream) -> None:	CODE
43 more matches not shown…

Modern AI Meta-Vocabulary27 hits · 87 pts

Severity	File	Line	Snippet	Context
MEDIUM	README.md	32	- [2026/05] 🔥 Agentic workload benchmark on AMD MI300X ([blog](https://blog.lmcache.ai/en/2026/05/12/benchmarking-lmcach	CODE
MEDIUM⚡	tests/v1/distributed/serde/test_asym_k16_v8_multi.py	34	# Test scaffolding (mirrors the _FakeMemoryObj in test_multi.py / test_fp8.py)	COMMENT
MEDIUM⚡	tests/v1/distributed/serde/test_multi.py	52	# Test scaffolding: a minimal MemoryObj stand-in mirroring test_fp8.py.	COMMENT
MEDIUM	docs/design/v1/gpu_connector/layout-invariant.md	42	├── detection.py # detect_format() orchestration	CODE
MEDIUM	…design/cli/commands/bench/engine_bench/bench-engine.md	42	# RAG answer quality (real QA data, documents cached individually)	COMMENT
MEDIUM	benchmarks/rag/README.md	1	# Benchmarking LLM Performance: RAG Use Case	COMMENT
MEDIUM	benchmarks/rag/README.md	16	## Running the RAG Benchmark	COMMENT
MEDIUM	lmcache/cli/commands/bench/server_bench/command.py	51	# orchestration safe.	COMMENT
MEDIUM	operator/AGENTS.md	355	Do NOT delete `// +kubebuilder:scaffold:*` comments. CLI injects code at these markers.	CODE
MEDIUM	operator/AGENTS.md	389	### Deploy Image Plugin (scaffold to deploy/manage ANY container image)	COMMENT
MEDIUM	operator/cmd/main.go	44	// +kubebuilder:scaffold:imports	COMMENT
MEDIUM	operator/cmd/main.go	57	// +kubebuilder:scaffold:scheme	COMMENT
MEDIUM	operator/cmd/main.go	175	// In the default scaffold provided, the program ends immediately after	COMMENT
MEDIUM	operator/cmd/main.go	221	// +kubebuilder:scaffold:builder	COMMENT
MEDIUM	operator/test/e2e/e2e_test.go	168	// +kubebuilder:scaffold:e2e-metrics-webhooks-readiness	COMMENT
MEDIUM	operator/test/e2e/e2e_test.go	208	// +kubebuilder:scaffold:e2e-webhooks-checks	COMMENT
MEDIUM	operator/config/crd/kustomization.yaml	8	# +kubebuilder:scaffold:crdkustomizeresource	COMMENT
MEDIUM	operator/config/crd/kustomization.yaml	13	# +kubebuilder:scaffold:crdkustomizewebhookpatch	COMMENT
MEDIUM⚡	operator/config/default/kustomization.yaml	225	# targets: # Do not remove or uncomment the following scaffold marker; required to generate code for target CRD.	COMMENT
MEDIUM⚡	operator/config/default/kustomization.yaml	226	# +kubebuilder:scaffold:crdkustomizecainjectionns	COMMENT
MEDIUM⚡	operator/config/default/kustomization.yaml	233	# targets: # Do not remove or uncomment the following scaffold marker; required to generate code for target CRD.	COMMENT
MEDIUM⚡	operator/config/default/kustomization.yaml	234	# +kubebuilder:scaffold:crdkustomizecainjectionname	COMMENT
MEDIUM	operator/config/samples/vllm_cacheblend_deployment.yaml	1	# Sample vLLM Deployment that OPTS IN to CacheBlend dependency injection.	COMMENT
MEDIUM	operator/config/samples/kustomization.yaml	6	# +kubebuilder:scaffold:manifestskustomizesamples	COMMENT
MEDIUM	operator/config/samples/vllm_lmcache_deployment.yaml	1	# Sample vLLM Deployment that OPTS IN to LMCacheEngine dependency injection.	COMMENT
MEDIUM	operator/internal/controller/suite_test.go	38	// +kubebuilder:scaffold:imports	COMMENT
MEDIUM	operator/internal/controller/suite_test.go	70	// +kubebuilder:scaffold:scheme	COMMENT

AI Slop Vocabulary24 hits · 74 pts

Severity	File	Line	Snippet	Context
MEDIUM⚡	tests/v1/test_mp_mem_kernels.py	31	# Tensor factories (ported from kernel harness)	COMMENT
MEDIUM	tests/v1/multiprocess/test_cache_server.py	193	#: backend has no such requirement, but this keeps the harness valid for	COMMENT
MEDIUM	tests/v1/multiprocess/test_engine_driven_transfer.py	275	# real IPC-backed factories to be usable in the harness.	COMMENT
MEDIUM	tests/cli/commands/trace/test_driver.py	272	# the test robust against future decorator additions.	COMMENT
MEDIUM	.github/workflows/operator_release.yml	51	# so behavior is robust whether or not that box is ticked in the UI.	COMMENT
MEDIUM	.buildkite/pipelines/comprehensive-tests.yml	9	- .buildkite/scripts/should-run-comprehensive.sh \|\| exit 0 # Skip comprehensive tests for safe-path-only PRs.	CODE
MEDIUM	.buildkite/pipelines/multiprocessing-test.yml	9	- .buildkite/scripts/should-run-comprehensive.sh \|\| exit 0 # Reuse the trigger condition for comprehensive tests	CODE
MEDIUM	.buildkite/k3_harness/setup-env.sh	287	# harness. Re-probe the full chain here so broken envs fail fast with the	COMMENT
MEDIUM	.buildkite/k3_harness/setup-blend-env.sh	17	# 2. Plugin-harness dep not in the LMCache/vLLM dependency closure.	COMMENT
MEDIUM	.buildkite/k3_harness/setup-blend-env.sh	59	# server-boot timeout inside the harness.	COMMENT
MEDIUM⚡	.buildkite/scripts/should-run-comprehensive.sh	4	# Decide whether comprehensive tests should run for this build.	COMMENT
MEDIUM⚡	.buildkite/scripts/should-run-comprehensive.sh	6	# 0 -> run comprehensive tests	COMMENT
MEDIUM⚡	.buildkite/scripts/should-run-comprehensive.sh	7	# 1 -> skip comprehensive tests (safe paths only)	COMMENT
MEDIUM⚡	.buildkite/scripts/should-run-comprehensive.sh	9	# If this isn't a PR build, always run comprehensive tests.	COMMENT
MEDIUM	.buildkite/scripts/should-run-comprehensive.sh	70	# Non-safe file touched -> run comprehensive tests.	COMMENT
MEDIUM	…ite/k3_tests/multiprocess/scripts/run-dsv4-flash-tp.sh	297	# across 128 tokens), making the comparison robust.	COMMENT
MEDIUM	…kite/k3_tests/multiprocess/scripts/launch-processes.sh	112	# robust to default changes (the default flipped False in #3869/#4437).	COMMENT
MEDIUM	…kite/k3_tests/multiprocess/scripts/launch-processes.sh	122	# rather than relying on the server default so the test is robust to default	COMMENT
MEDIUM	…dkite/k3_tests/multiprocess/scripts/run-single-test.sh	9	# This mirrors the comprehensive tests' run-single-config.sh pattern.	COMMENT
MEDIUM	…ite/k3_tests/comprehensive/scripts/upload-baselines.sh	3	# Runs AFTER all comprehensive config steps complete (via `wait`).	COMMENT
LOW	lmcache/v1/memory_management.py	1047	# so just return None here	COMMENT
LOW	…cache/v1/memory_allocators/cu_file_memory_allocator.py	21	# layer - for now just use the current device.	COMMENT
MEDIUM	lmcache/v1/gpu_connector/utils.py	775	# Also accept [P,2,B,H,D] (or [NB,2,BS,NH,HS]) to be more robust.	COMMENT
MEDIUM	lmcache/cli/commands/bench/l2_adapter_bench/command.py	457	# the one-time tensor allocation + fill (benchmark harness work,	COMMENT

Cross-Language Confusion8 hits · 35 pts

Severity	File	Line	Snippet	Context
HIGH	…s/v1/multiprocess/test_lmcache_driven_transfer_skip.py	57	# chunk size = 2 blocks. Chunk 0 = [0, 0] (null), chunk 1 = [0, 9] (has a	COMMENT
HIGH	…s/v1/multiprocess/test_lmcache_driven_transfer_skip.py	89	# chunk 0: kg0=0 and kg1=0 -> null; chunk 1: kg0=0 but kg1=4 -> not null.	COMMENT
HIGH	tests/v1/shm_allocator/verify_shmfile_tcp.py	15	mkdir -p tests/v1/shm_allocator/csrc/build && cd tests/v1/shm_allocator/csrc/build	STRING
HIGH	lmcache/v1/cache_controller/worker.py	643	self.push(),	CODE
HIGH	lmcache/v1/storage_backend/nixl_storage_backend.py	1352	self.pool.push(metadata.index)	CODE
HIGH	lmcache/v1/internal_api_server/vllm/cache_api.py	346	# "locations": null, "request_configs": null}	STRING
HIGH	lmcache/v1/internal_api_server/vllm/cache_api.py	354	# "request_configs": null}	STRING
HIGH	…ommands/bench/engine_bench/workloads/rag_qa_quality.py	527	An unparsed sample's ``f1`` is null, not zero, so a cross-run diff can	STRING

Hallucination Indicators2 hits · 20 pts

Severity	File	Line	Snippet	Context
CRITICAL	tests/v1/mp_coordinator/test_directory_api.py	206	key_directory = client.app.state.ctx.views.get(KeyDirectory)	CODE
CRITICAL	lmcache/v1/storage_backend/connector/azure_connector.py	145	from azure.core.exceptions import ResourceNotFoundError	CODE

AI Structural Patterns20 hits · 17 pts

Severity	File	Line	Context
LOW	tests/v1/mp_coordinator/test_key_directory.py	26	CODE
LOW	tests/v1/mp_coordinator/test_event_gate.py	37	CODE
LOW	tests/cli/commands/bench/engine_bench/test_stats.py	25	CODE
LOW	…ds/bench/engine_bench/workloads/test_rag_qa_quality.py	70	CODE
LOW	…che/v1/distributed/l2_adapters/aerospike_l2_adapter.py	48	CODE
LOW	…che/v1/distributed/l2_adapters/raw_block_l2_adapter.py	216	CODE
LOW	…1/distributed/l2_adapters/sagemaker_hyperpod_client.py	159	CODE
LOW	…stributed/l2_adapters/sagemaker_hyperpod_l2_adapter.py	60	CODE
LOW	…ache/v1/distributed/l2_adapters/bigtable_l2_adapter.py	172	CODE
LOW	lmcache/v1/distributed/l2_adapters/valkey_l2_adapter.py	165	CODE
LOW	lmcache/v1/storage_backend/valkey/worker_pool.py	173	CODE
LOW	…/v1/storage_backend/native_clients/aerospike_client.py	22	CODE
LOW	lmcache/v1/storage_backend/connector/eic_connector.py	423	CODE
LOW	…cache/v1/storage_backend/connector/valkey_connector.py	107	CODE
LOW	lmcache/v1/kv_codec/asym_k16_v8.py	271	CODE
LOW	lmcache/cli/commands/bench/server_bench/helpers.py	308	CODE
LOW	lmcache/cli/commands/bench/server_bench/helpers.py	426	CODE
LOW	lmcache/cli/commands/bench/server_bench/helpers.py	823	CODE
LOW	lmcache/cli/commands/bench/server_bench/helpers.py	892	CODE
LOW	lmcache/cli/commands/bench/server_bench/helpers.py	1069	CODE

Example Usage Blocks8 hits · 13 pts

Severity	File	Line	Snippet	Context
LOW⚡	tests/v1/shm_allocator/shmfile_basic_check.yaml	3	# Usage:	COMMENT
LOW	…es/lmc_external_l2_adapter/scripts/install_and_test.sh	7	# Usage:	COMMENT
LOW	.github/scripts/download_model.sh	8	# Usage:	COMMENT
LOW	.github/scripts/install_lmcache_cpu.sh	9	# Usage:	COMMENT
LOW	.github/scripts/install_vllm_cpu.sh	16	# Usage:	COMMENT
LOW	.buildkite/k3_harness/install-agent-stack.sh	4	# Usage:	COMMENT
LOW	.buildkite/k3_harness/resolve-pinned-vllm.sh	25	# Usage:	COMMENT
LOW	.buildkite/k3_tests/common_scripts/path-filter.sh	5	# Usage:	COMMENT

TODO Padding4 hits · 6 pts

Severity	File	Line	Snippet	Context
LOW	lmcache/v1/memory_management.py	855	# TODO: consider using one of the alternatives	COMMENT
LOW	lmcache/v1/multiprocess/mq.py	489	# TODO: implement this in the future versions if needed	COMMENT
LOW	lmcache/v1/storage_backend/gds_backend.py	532	# TODO: implement pin() semantics	COMMENT
LOW	lmcache/cli/commands/query/kvcache_command.py	24	# TODO: implement kvcache query logic	COMMENT

Fake / Example Data3 hits · 3 pts

Severity	File	Line	Snippet	Context
LOW	tests/v1/test_sglang_mp_adapter.py	342	conn._create_key = lambda args, *kwargs: "fake-key" # type: ignore[method-assign,assignment]	CODE
LOW	benchmarks/multi_doc_qa/shuffle_doc_qa.py	263	api_key="dummy-key",	CODE
LOW	benchmarks/ttft-estimator/ttft-estimator.py	37	client = OpenAI(api_key="dummy-key", base_url="http://localhost:8000/v1")	CODE

Slop Phrases1 hit · 3 pts

Severity	File	Line	Snippet	Context
MEDIUM	.github/workflows/codeql.yml	59	# If you are analyzing a compiled language, you can modify the 'build-mode' for that language to customize how	COMMENT

Overly Generic Function Names2 hits · 2 pts

Severity	File	Line	Snippet	Context
LOW	tests/v1/mp_coordinator/test_cache_events.py	694	def handle_request(self, request: httpx.Request) -> httpx.Response:	CODE
LOW	lmcache/v1/cache_controller/worker.py	529	async def handle_request(self):	CODE

Dead Code1 hit · 2 pts

Severity	File	Line	Snippet	Context
MEDIUM	lmcache/v1/mp_coordinator/persistence/store.py	122		CODE

Analysis Overview

What These Metrics Mean

Score History

Severity Breakdown

Directory Score Breakdown

Pattern Findings