simonw/llm

11.9

Adjusted Score

11.9

Raw Score

100%

Time Factor

2026-07-09

Last Push

12.2K

Stars

Python

Language

40.0K

Lines of Code

106

Files

490

Pattern Hits

2026-07-14

Scan Date

0.05

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 0HIGH 5MEDIUM 8LOW 477

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 490 distinct pattern matches across 12 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.

Hyper-Verbose Identifiers391 hits · 348 pts

Severity	File	Line	Snippet	Context
LOW⚡	llm/hookspecs.py	19	def register_embedding_models(register):	CODE
LOW⚡	llm/hookspecs.py	24	def register_template_loaders(register):	CODE
LOW⚡	llm/hookspecs.py	29	def register_fragment_loaders(register):	CODE
LOW	llm/models.py	187	def _get_arguments_input_schema(function, name):	CODE
LOW	llm/models.py	216	def _implementation_arguments(tool: "Tool", tool_call: "ToolCall") -> dict:	CODE
LOW	llm/models.py	2632	def _append_tool_results_to_chain(chain, tool_results, attachments) -> List[Any]:	CODE
LOW	llm/models.py	2684	def _trailing_pending_tool_calls(messages) -> List[ToolCall]:	CODE
LOW	llm/models.py	3441	def _remove_titles_recursively(obj):	CODE
LOW	llm/__init__.py	234	def get_embedding_models_with_aliases() -> List["EmbeddingModelWithAliases"]:	CODE
LOW	llm/__init__.py	276	def get_embedding_model_aliases() -> Dict[str, EmbeddingModel]:	CODE
LOW	llm/__init__.py	498	def get_default_embedding_model():	CODE
LOW	llm/__init__.py	502	def set_default_embedding_model(model):	CODE
LOW	llm/embeddings.py	173	def embed_multi_with_metadata(	CODE
LOW	llm/cli.py	111	async def display_async_stream_events(events, *, show_reasoning=True):	CODE
LOW	llm/cli.py	206	def process_fragments_in_chat(	CODE
LOW	llm/cli.py	297	def resolve_attachment_with_type(value: str, mimetype: str) -> Attachment:	CODE
LOW	llm/cli.py	313	def attachment_types_callback(ctx, param, values) -> List[Attachment]:	CODE
LOW	llm/cli.py	2072	def _format_tool_call_arguments(arguments):	STRING
LOW	llm/cli.py	2349	def model_matches_id_or_alias(model_with_aliases, model_ids):	STRING
LOW	llm/cli.py	2356	def render_model_with_aliases(	STRING
LOW	llm/cli.py	2424	def render_model_with_options(model_id, *, async_=False):	STRING
LOW	llm/utils.py	182	def extract_fenced_code_block(text: str, last: bool = False) -> Optional[str]:	CODE
LOW	llm/default_plugins/openai_models.py	382	def register_embedding_models(register):	CODE
LOW	llm/default_plugins/openai_models.py	1409	def _reasoning_text_from_item(self, item):	CODE
LOW	tests/test_cli_openai_models.py	46	def test_openai_options_min_max():	CODE
LOW	tests/test_cli_openai_models.py	205	def test_openai_image_detail_option_description(model_id, expected_description):	CODE
LOW	tests/test_cli_openai_models.py	210	def test_openai_image_detail_option_is_sent_on_image_attachments(httpx_mock):	CODE
LOW	tests/test_cli_openai_models.py	357	def test_openai_image_detail_original_is_rejected_for_other_models():	CODE
LOW	tests/test_async_parity.py	16	def test_echo_registered_for_both():	CODE
LOW	tests/test_async_parity.py	25	async def test_async_to_dict_captures_chain_and_output():	CODE
LOW	tests/test_async_parity.py	39	async def test_async_to_dict_raises_before_awaited():	CODE
LOW	tests/test_async_parity.py	47	async def test_async_from_dict_rehydrates():	CODE
LOW	tests/test_async_parity.py	65	async def test_async_from_dict_then_reply_continues():	CODE
LOW	tests/test_async_parity.py	89	async def test_async_from_row_response_messages_synthesized(tmp_path):	CODE
LOW	tests/test_async_parity.py	121	async def test_async_load_conversation_follow_up_preserves_chain(tmp_path):	CODE
LOW	tests/test_async_parity.py	172	async def test_sync_and_async_produce_identical_chain():	CODE
LOW	tests/test_async_parity.py	191	async def test_async_chain_tool_result_turn_has_full_chain():	CODE
LOW	tests/test_async_parity.py	230	async def test_astream_events_matches_stream_events_for_text_only():	CODE
LOW	tests/test_async_parity.py	260	async def test_async_from_dict_model_override():	CODE
LOW	tests/test_async_parity.py	272	def test_sync_from_dict_model_override():	CODE
LOW	tests/test_async_parity.py	284	async def test_async_to_dict_preserves_datetime():	CODE
LOW	tests/test_async_parity.py	294	async def test_async_to_dict_preserves_usage_when_set(async_mock_model):	CODE
LOW	tests/test_async_parity.py	312	async def test_async_reply_messages_kwarg_appends():	CODE
LOW	tests/test_async_parity.py	325	async def test_async_full_chain_to_dict_round_trip_three_turns():	CODE
LOW	tests/test_async_parity.py	365	async def test_async_reply_chains_three_turns():	CODE
LOW	tests/test_utils.py	112	def test_extract_fenced_code_block(input, last, expected):	CODE
LOW	tests/test_utils.py	318	def test_test_truncate_string_keep_end(	CODE
LOW	tests/test_utils.py	386	def test_backtick_count_adjustment(content: str, backtick_count: int):	CODE
LOW	tests/test_utils.py	471	def test_toolbox_config_capture():	CODE
LOW	tests/conftest.py	44	def user_path_with_embeddings(user_path):	CODE
LOW	tests/conftest.py	212	def register_embed_demo_model(embed_demo, mock_model, async_mock_model):	CODE
LOW	tests/conftest.py	217	def register_embedding_models(self, register):	CODE
LOW	tests/conftest.py	265	def mocked_openai_chat_returning_fenced_code(httpx_mock):	CODE
LOW	tests/conftest.py	309	def mocked_openai_chat_stream(httpx_mock):	CODE
LOW	tests/conftest.py	415	def mocked_openai_completion_logprobs_stream(httpx_mock):	CODE
LOW	tests/conftest.py	426	def mocked_openai_completion_logprobs(httpx_mock):	CODE
LOW	tests/test_llm_logs.py	138	def test_logs_text_with_options(user_path):	CODE
LOW	tests/test_llm_logs.py	174	def test_logs_token_usage_details_are_markdown_code(user_path):	CODE
LOW	tests/test_llm_logs.py	252	def test_logs_extract_first_code(args, log_path):	CODE
LOW	tests/test_llm_logs.py	271	def test_logs_extract_last_code(args, log_path):	CODE
331 more matches not shown…

Cross-Language Confusion5 hits · 30 pts

Severity	File	Line	Snippet	Context
HIGH	llm/utils.py	668	if re.match(r'\s*(["\[\d\-]\|true\|false\|null)', arg_body, re.I):	CODE
HIGH	tests/test_llm_logs.py	1179	"""When reasoning is empty/null, no heading appears (existing	STRING
HIGH⚡	tests/test_plugins.py	467	'[{"id": 2, "tool_id": 1, "name": "upper", "output": "ONE", "tool_call_id": "tc_TCID", "exception": null	CODE
HIGH⚡	tests/test_plugins.py	472	'[{"id": 3, "tool_id": 1, "name": "upper", "output": "TWO", "tool_call_id": "tc_TCID", "exception": null	CODE
HIGH⚡	tests/test_plugins.py	480	'[{"id": 4, "tool_id": 1, "name": "upper", "output": "THREE", "tool_call_id": "tc_TCID", "exception": nu	CODE

Deep Nesting27 hits · 26 pts

Severity	File	Line	Context
LOW	llm/plugins.py	21	CODE
LOW	llm/models.py	520	CODE
LOW	llm/models.py	1112	CODE
LOW	llm/models.py	1415	CODE
LOW	llm/models.py	1833	CODE
LOW	llm/models.py	2173	CODE
LOW	llm/cli.py	131	CODE
LOW	llm/cli.py	519	CODE
LOW	llm/cli.py	1086	CODE
LOW	llm/cli.py	1657	CODE
LOW	llm/cli.py	2743	CODE
LOW	llm/cli.py	3229	CODE
LOW	llm/cli.py	3359	CODE
LOW	llm/cli.py	3554	CODE
LOW	llm/cli.py	3454	CODE
LOW	llm/utils.py	77	CODE
LOW	llm/utils.py	312	CODE
LOW	llm/utils.py	554	CODE
LOW	llm/parts.py	253	CODE
LOW	llm/default_plugins/openai_models.py	2029	CODE
LOW	llm/default_plugins/openai_models.py	775	CODE
LOW	llm/default_plugins/openai_models.py	942	CODE
LOW	llm/default_plugins/openai_models.py	1060	CODE
LOW	llm/default_plugins/openai_models.py	1223	CODE
LOW	llm/default_plugins/openai_models.py	1511	CODE
LOW	llm/default_plugins/openai_models.py	1738	CODE
LOW	tests/test_templates.py	464	CODE

Unused Imports24 hits · 19 pts

Severity	File	Line	Context
LOW	llm/__init__.py	1	CODE
LOW	llm/__init__.py	2	CODE
LOW	llm/__init__.py	2	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	6	CODE
LOW	llm/__init__.py	30	CODE
LOW	llm/__init__.py	30	CODE
LOW	llm/__init__.py	30	CODE
LOW	llm/__init__.py	30	CODE
LOW	llm/__init__.py	30	CODE
LOW	llm/__init__.py	37	CODE
LOW	llm/__init__.py	38	CODE

Excessive Try-Catch Wrapping14 hits · 14 pts

Severity	File	Line	Snippet	Context
LOW	llm/models.py	1927	except Exception as ex:	CODE
LOW	llm/models.py	2240	except Exception as ex:	CODE
LOW	llm/models.py	2295	except Exception as ex:	CODE
LOW	llm/models.py	2339	except Exception as ex:	CODE
LOW	llm/models.py	2363	except Exception as ex:	CODE
LOW	llm/models.py	2381	except Exception as ex:	CODE
MEDIUM	llm/cli.py	290	def convert(self, value, param, ctx):	CODE
LOW	llm/cli.py	187	except Exception as ex:	CODE
LOW	llm/cli.py	961	except Exception as ex:	CODE
LOW	llm/cli.py	1477	except Exception as ex:	CODE
LOW	llm/cli.py	4083	except Exception as ex:	STRING
LOW	llm/default_plugins/openai_models.py	1445	except Exception:	CODE
LOW	docs/plugins/plugin-hooks.md	221	except Exception as e:	CODE
LOW	docs/plugins/plugin-hooks.md	271	except Exception as ex:	CODE

AI Structural Patterns12 hits · 10 pts

Severity	File	Line	Context
LOW	llm/models.py	411	CODE
LOW	llm/models.py	650	CODE
LOW	llm/models.py	700	CODE
LOW	llm/models.py	778	CODE
LOW	llm/models.py	831	CODE
LOW	llm/models.py	3105	CODE
LOW	llm/models.py	3145	CODE
LOW	llm/models.py	3224	CODE
LOW	llm/models.py	3264	CODE
LOW	llm/default_plugins/openai_models.py	710	CODE
LOW	llm/default_plugins/openai_models.py	1460	CODE
LOW	llm/default_plugins/openai_models.py	1689	CODE

Self-Referential Comments4 hits · 10 pts

Severity	File	Line	Snippet	Context
MEDIUM	llm/cli.py	3833	# Create a test Options object to validate	STRING
MEDIUM	tests/test_templates.py	158	# Create a file to test attachment	COMMENT
MEDIUM	tests/test_chat_templates.py	35	# Create a system fragment file	COMMENT
MEDIUM	docs/conf.py	6	# This file is execfile()d with the current directory set to its	COMMENT

Modern AI Meta-Vocabulary3 hits · 8 pts

Severity	File	Line	Snippet	Context
MEDIUM	llm/__init__.py	439	# Try to resolve it to an embedding model	COMMENT
MEDIUM	tests/test_openai_responses.py	574	# item, otherwise the model loses its chain of thought.	COMMENT
MEDIUM	docs/embeddings/python-api.md	40	To work with embeddings in this way you will need an instance of a [sqlite-utils Database](https://sqlite-utils.datasett	CODE

Redundant / Tautological Comments4 hits · 6 pts

Severity	File	Line	Snippet	Context
LOW	llm/cli.py	3495	# Read the file twice - first time is to get a count	COMMENT
LOW	llm/utils.py	377	# Check if the schema is newline-separated or comma-separated	COMMENT
LOW	tests/test_utils.py	392	# Check if the correct number of backticks is used	COMMENT
LOW	.github/workflows/stable-docs.yml	52	# Check if the list of files is non-empty	COMMENT

Modern Structural Boilerplate3 hits · 2 pts

Severity	File	Line	Snippet	Context
LOW	llm/__init__.py	49	__all__ = [	CODE
LOW	llm/cli.py	3986	def set_model_option(model_id: str, key: str, value: Any) -> None:	STRING
LOW	llm/serialization.py	41	__all__ = [	CODE

Over-Commented Block2 hits · 2 pts

Severity	File	Line	Snippet	Context
LOW	docs/conf.py	1	#!/usr/bin/env python3	COMMENT
LOW	docs/conf.py	121	# -- Options for LaTeX output ---------------------------------------------	COMMENT

Overly Generic Function Names1 hit · 1 pts

Severity	File	Line	Snippet	Context
LOW	llm/utils.py	108	def handle_request(self, request: httpx.Request) -> httpx.Response:	CODE

Analysis Overview

What These Metrics Mean

Score History

Severity Breakdown

Directory Score Breakdown

Pattern Findings