AI Slop Code RIsk Monitor

StarTrail-org/PixelRAG

https://arxiv.org/abs/2606.28344. The end of web parsing. The beginning of scalable pixel-native search. link: https://pixelrag.ai/

26.5 Elevated pattern signal View on GitHub

Maintenance and Governance Review

This report summarizes scanner evidence for StarTrail-org/PixelRAG, a Python project with 10,249 GitHub stars. SynthScan examined 78,957 lines of code across 309 source files, recording 1207 pattern matches distributed across 21 syntactic categories. The legacy adjusted pattern signal is 26.5 (Elevated pattern signal); it is not an authorship estimate.

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

26.5
Adjusted Signal
26.5
Raw Pattern Score
100%
Time Factor
2026-10-08
Last Push
10.2K
Stars
Python
Language
79.0K
Lines of Code
309
Files
1.2K
Pattern Hits
2026-10-12
Scan Date
0.39
HC Hit Rate

Risk Review · v3

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

Cognitive Ownership Risk 28.0/100 · MEDIUM

Can maintainers explain intent, assumptions, and trade-offs? 430 finding(s).

Hallucinated Dependency & Supply Chain Risk 14.7/100 · LOW

Are imports declared and package/API names verified? 167 finding(s).

Agent Over-Engineering Ratio 61.8/100 · HIGH

Are wrappers, guards, and repeated structures necessary? 804 finding(s).

PR Review Velocity Impact 76.3/100 · CRITICAL

Which changes need focused human review? 1207 finding(s).

What These Metrics Mean

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

Pattern-Signal History

Run additional scans to build a history of the legacy pattern signal. Score changes are not a measure of AI adoption or code quality.

No multi-scan history yet — run the scanner again to build trend data.

Severity Breakdown

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

CRITICAL 0HIGH 121MEDIUM 232LOW 854

Directory Score Breakdown

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

Pattern Findings

The scanner identified 1207 distinct pattern matches across 21 syntactic categories. Each entry below represents a discrete location in the source code where the engine recorded a detector match to inspect in context. Expand any category row to inspect the individual file paths, line numbers, code snippets, and the lexical context (CODE, COMMENT, or STRING) in which each match was detected.

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

Decorative Section Separators204 hits · 616 pts
SeverityFileLineSnippetContext
MEDIUM⚡embed/src/pixelrag_embed/embed.py747# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡embed/src/pixelrag_embed/embed.py749# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡embed/src/pixelrag_embed/embed.py758# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡embed/src/pixelrag_embed/embed.py760# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡embed/src/pixelrag_embed/embed.py1822# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡embed/src/pixelrag_embed/embed.py1824# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py156# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py158# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py180# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py182# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py215# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py217# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py397# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py399# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py486# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py488# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py532# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py534# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py594# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py596# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py2201# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py2203# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py2227# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py2229# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py2629# ---------------------------------------------------------------------------COMMENT
MEDIUMembed/src/pixelrag_embed/embed.py2631# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡tests/test_department_filter.py14# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡tests/test_department_filter.py16# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡tests/test_department_filter.py54# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡tests/test_department_filter.py56# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡render/src/pixelrag_render/bench/bench_throughput.py232# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡render/src/pixelrag_render/bench/bench_throughput.py234# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡render/src/pixelrag_render/bench/bench_throughput.py382# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡render/src/pixelrag_render/bench/bench_throughput.py384# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/bench/bench_throughput.py46# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/bench/bench_throughput.py48# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/bench/bench_throughput.py127# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/bench/bench_throughput.py129# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/bench/bench_throughput.py267# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/bench/bench_throughput.py269# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/bench/bench_throughput.py489# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/bench/bench_throughput.py491# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/strategies/connection.py49 # ------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/strategies/connection.py51 # ------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/strategies/connection.py103 # ------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/strategies/connection.py105 # ------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/backends/fast_cdp.py97# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/backends/fast_cdp.py99# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/backends/fast_cdp.py157# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/backends/fast_cdp.py159# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/backends/fast_cdp.py310# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/backends/fast_cdp.py312# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/backends/fast_cdp.py681# ---------------------------------------------------------------------------COMMENT
MEDIUMrender/src/pixelrag_render/backends/fast_cdp.py683# ---------------------------------------------------------------------------COMMENT
MEDIUMtrain/generate_query_pairs.py65# ── Page filtering ───────────────────────────────────────────────────COMMENT
MEDIUM⚡train/train_contrastors.py67# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡train/train_contrastors.py69# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡train/train_contrastors.py952# ---------------------------------------------------------------------------COMMENT
MEDIUM⚡train/train_contrastors.py954# ---------------------------------------------------------------------------COMMENT
MEDIUMtrain/train_contrastors.py544# ---------------------------------------------------------------------------COMMENT
144 more matches not shown…
Empty Error Handlers64 hits · 320 pts
SeverityFileLineSnippetContext
HIGHembed/src/pixelrag_embed/embed.py1671except Exception:CODE
HIGHembed/src/pixelrag_embed/embed.py2178except OSError:CODE
HIGHembed/src/pixelrag_embed/embed.py1636except Exception:CODE
HIGHembed/src/pixelrag_embed/embed.py1663except Exception:CODE
HIGHembed/src/pixelrag_embed/embed.py2524except OSError:CODE
HIGHembed/src/pixelrag_embed/embed.py355except (json.JSONDecodeError, OSError):CODE
HIGHembed/src/pixelrag_embed/embed.py1348except Exception:CODE
HIGHembed/src/pixelrag_embed/embed_cpu.py95except (json.JSONDecodeError, OSError):CODE
HIGHrender/src/pixelrag_render/chrome.py146except Exception:CODE
HIGHrender/src/pixelrag_render/bench/bench_throughput.py123except OSError:CODE
HIGHrender/src/pixelrag_render/bench/bench_throughput.py335except OSError:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_oneshot.py121except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_oneshot.py128except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_oneshot.py336except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_dc_single.py178except Exception:CODE
HIGH…r/src/pixelrag_render/strategies/cdp_pipelined_tabs.py192except OSError:CODE
HIGH…r/src/pixelrag_render/strategies/cdp_pipelined_tabs.py181except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_sequential.py208except Exception:CODE
HIGH…/src/pixelrag_render/strategies/cdp_pertile_imgwait.py182except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/connection.py162except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/connection.py187except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/connection.py191except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/connection.py219except OSError:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_phased.py238except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_phased.py280except Exception:CODE
HIGH…der/src/pixelrag_render/strategies/cdp_pipelined_dc.py66except Exception:CODE
HIGH…der/src/pixelrag_render/strategies/cdp_pipelined_dc.py70except Exception:CODE
HIGH…der/src/pixelrag_render/strategies/cdp_pipelined_dc.py294except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_overlap.py187except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_overlap.py191except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_overlap.py300except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_overlap.py201except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_multitab.py190except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_multitab.py303except Exception:CODE
HIGHrender/src/pixelrag_render/strategies/cdp_multitab.py200except Exception:CODE
HIGHrender/src/pixelrag_render/backends/cdp.py327except (KeyError, TypeError):CODE
HIGHrender/src/pixelrag_render/backends/cdp.py476except Exception:CODE
HIGHrender/src/pixelrag_render/backends/cdp.py847except Exception:CODE
HIGHrender/src/pixelrag_render/backends/fast_cdp.py110except (OSError, AttributeError):CODE
HIGHrender/src/pixelrag_render/backends/fast_cdp.py299except Exception:CODE
HIGHrender/src/pixelrag_render/backends/fast_cdp.py653except Exception:CODE
HIGHrender/src/pixelrag_render/backends/fast_cdp.py538except Exception:CODE
HIGHtrain/fake_data.py55except (json.JSONDecodeError, KeyError):CODE
HIGHtrain/prepare_hf_dataset.py88except OSError:CODE
HIGHtrain/prepare_hf_dataset_single_jsonl.py85except OSError:CODE
HIGHtrain/clean_queries_simpleqa_style.py177except json.JSONDecodeError:CODE
HIGHtrain/filter_self_contained.py168except Exception:CODE
HIGHtrain/sft/download_tiles.py229except Exception:CODE
HIGHindex/src/pixelrag_index/monitor.py105except ImportError:CODE
HIGHindex/src/pixelrag_index/monitor.py193except (subprocess.TimeoutExpired, Exception):CODE
HIGHindex/src/pixelrag_index/monitor.py239except (subprocess.TimeoutExpired, Exception):CODE
HIGHindex/src/pixelrag_index/monitor.py322except (subprocess.TimeoutExpired, Exception):CODE
HIGHindex/src/pixelrag_index/monitor.py648except (ValueError, TypeError):CODE
HIGHindex/src/pixelrag_index/sources/kiwix.py243except (OSError, ProcessLookupError):CODE
HIGHindex/src/pixelrag_index/sources/kiwix.py623except Exception:CODE
HIGHindex/src/pixelrag_index/sources/kiwix.py257except (OSError, ProcessLookupError):CODE
HIGHeval/lib/retrieval.py3733except Exception:CODE
HIGHeval/lib/retrieval.py3566except ValueError:CODE
HIGHeval/lib/simpleqa_data.py938except (ValueError, SyntaxError):CODE
HIGHeval/lib/simpleqa_data.py231except (ValueError, SyntaxError):CODE
4 more matches not shown…
Excessive Try-Catch Wrapping295 hits · 270 pts
SeverityFileLineSnippetContext
LOW⚡embed/src/pixelrag_embed/embed.py1663 except Exception:CODE
LOW⚡embed/src/pixelrag_embed/embed.py1671 except Exception:CODE
LOWembed/src/pixelrag_embed/embed.py104 except Exception:CODE
LOWembed/src/pixelrag_embed/embed.py417 except Exception as e:CODE
LOWembed/src/pixelrag_embed/embed.py432 except Exception as e:CODE
LOWembed/src/pixelrag_embed/embed.py837 except Exception as e:CODE
LOWembed/src/pixelrag_embed/embed.py928 except Exception as e:CODE
LOWembed/src/pixelrag_embed/embed.py1000 except Exception as exc:CODE
LOWembed/src/pixelrag_embed/embed.py1051 except Exception as e:CODE
LOWembed/src/pixelrag_embed/embed.py1080 except Exception as e:CODE
LOWembed/src/pixelrag_embed/embed.py1230 except Exception:CODE
LOWembed/src/pixelrag_embed/embed.py1324 except Exception:CODE
LOWembed/src/pixelrag_embed/embed.py1333 except Exception:CODE
LOWembed/src/pixelrag_embed/embed.py1348 except Exception:CODE
LOWembed/src/pixelrag_embed/embed.py1636 except Exception:CODE
LOWembed/src/pixelrag_embed/embed.py1945 except Exception as e:CODE
LOWembed/src/pixelrag_embed/embed.py2214 except Exception as e:CODE
LOWembed/src/pixelrag_embed/chunk.py150 except Exception as e:CODE
LOWembed/src/pixelrag_embed/chunk.py474 except Exception as e:CODE
LOWrender/src/pixelrag_render/render.py388 except Exception as e:CODE
LOWrender/src/pixelrag_render/chrome.py146 except Exception:CODE
LOWrender/src/pixelrag_render/chrome.py166 except Exception:CODE
LOW⚡render/src/pixelrag_render/bench/bench_throughput.py225 except Exception as e:CODE
LOWrender/src/pixelrag_render/bench/bench_throughput.py112 except Exception:CODE
LOWrender/src/pixelrag_render/bench/bench_throughput.py248 except Exception:CODE
LOWrender/src/pixelrag_render/bench/bench_throughput.py451 except Exception as e:CODE
LOWrender/src/pixelrag_render/strategies/cdp_oneshot.py107 except Exception as e:CODE
LOW⚡render/src/pixelrag_render/strategies/cdp_oneshot.py121 except Exception:CODE
LOW⚡render/src/pixelrag_render/strategies/cdp_oneshot.py125 except Exception:CODE
LOW⚡render/src/pixelrag_render/strategies/cdp_oneshot.py128 except Exception:CODE
LOWrender/src/pixelrag_render/strategies/cdp_oneshot.py213 except Exception as e:CODE
LOWrender/src/pixelrag_render/strategies/cdp_oneshot.py226 except Exception:CODE
LOWrender/src/pixelrag_render/strategies/cdp_oneshot.py249 except Exception as e:CODE
LOWrender/src/pixelrag_render/strategies/cdp_oneshot.py264 except Exception as e:CODE
LOWrender/src/pixelrag_render/strategies/cdp_oneshot.py274 except Exception as e:CODE
LOWrender/src/pixelrag_render/strategies/cdp_oneshot.py292 except Exception:CODE
LOWrender/src/pixelrag_render/strategies/cdp_oneshot.py336 except Exception:CODE
LOWrender/src/pixelrag_render/strategies/cdp_oneshot.py366 except Exception as e:CODE
LOWrender/src/pixelrag_render/strategies/cdp_fullpage.py114 except Exception as e:CODE
LOW⚡render/src/pixelrag_render/strategies/cdp_fullpage.py136 except Exception:CODE
LOW⚡render/src/pixelrag_render/strategies/cdp_fullpage.py191 except Exception as e:CODE
LOWrender/src/pixelrag_render/strategies/cdp_dc_single.py121 except Exception as e:CODE
LOWrender/src/pixelrag_render/strategies/cdp_dc_single.py135 except Exception:CODE
LOWrender/src/pixelrag_render/strategies/cdp_dc_single.py178 except Exception:CODE
LOWrender/src/pixelrag_render/strategies/cdp_dc_single.py205 except Exception as e:CODE
LOW…r/src/pixelrag_render/strategies/cdp_pipelined_tabs.py168 except Exception:CODE
LOW…r/src/pixelrag_render/strategies/cdp_pipelined_tabs.py181 except Exception:CODE
LOW…r/src/pixelrag_render/strategies/cdp_pipelined_tabs.py247 except Exception:CODE
LOW…r/src/pixelrag_render/strategies/cdp_pipelined_tabs.py300 except Exception as e:CODE
LOWrender/src/pixelrag_render/strategies/cdp_noscroll.py107 except Exception as e:CODE
LOWrender/src/pixelrag_render/strategies/cdp_noscroll.py119 except Exception:CODE
LOWrender/src/pixelrag_render/strategies/cdp_noscroll.py173 except Exception as e:CODE
LOWrender/src/pixelrag_render/strategies/cdp_sequential.py125 except Exception as e:CODE
LOWrender/src/pixelrag_render/strategies/cdp_sequential.py163 except Exception:CODE
LOWrender/src/pixelrag_render/strategies/cdp_sequential.py208 except Exception:CODE
LOWrender/src/pixelrag_render/strategies/cdp_sequential.py235 except Exception as e:CODE
LOW…/src/pixelrag_render/strategies/cdp_pertile_imgwait.py123 except Exception as e:CODE
LOW…/src/pixelrag_render/strategies/cdp_pertile_imgwait.py138 except Exception:CODE
LOW…/src/pixelrag_render/strategies/cdp_pertile_imgwait.py182 except Exception:CODE
LOW…/src/pixelrag_render/strategies/cdp_pertile_imgwait.py209 except Exception as e:CODE
235 more matches not shown…
Hyper-Verbose Identifiers179 hits · 190 pts
SeverityFileLineSnippetContext
LOW⚡embed/src/pixelrag_embed/embed.py1667def _close_all_persistent_pools() -> None:CODE
LOWembed/src/pixelrag_embed/embed.py437def read_tiles_and_hash_parallel(CODE
LOWembed/src/pixelrag_embed/embed.py694def _get_tokenizer_direct_gpu(engine):CODE
LOWembed/src/pixelrag_embed/embed.py889def _embed_tile_infos_with_engine(CODE
LOWembed/src/pixelrag_embed/embed.py1240def _gpu_worker_persistent_entry(CODE
LOWtests/test_serve_ondemand_budget.py96def test_renders_are_capped_per_request(client):CODE
LOWtests/test_serve_ondemand_budget.py104def test_budget_spans_all_queries_in_one_request(client):CODE
LOWtests/test_serve_ondemand_budget.py120def test_cache_hits_do_not_spend_the_budget(client):CODE
LOW⚡tests/test_pdf_manifest.py42def test_pdf_manifest_reports_a_full_render_as_complete(three_page_pdf, tmp_path):CODE
LOW⚡tests/test_pdf_manifest.py51def test_pdf_manifest_reports_a_page_subset_as_incomplete(three_page_pdf, tmp_path):CODE
LOW⚡tests/test_pdf_manifest.py61def test_pdf_manifest_records_a_full_page_range_as_incomplete(three_page_pdf, tmp_path):CODE
LOWtests/test_serve_reconstruct.py15def test_vector_ids_are_capped():CODE
LOWtests/test_bug_fixes.py19def test_load_config_deep_merge(tmp_path):CODE
LOWtests/test_bug_fixes.py38def test_local_source_generates_valid_as_uri(tmp_path):CODE
LOWtests/test_bug_fixes.py56def test_department_of_resolves_file_uri(tmp_path):CODE
LOWtests/test_bug_fixes.py69def test_pixelshot_forwards_extract_text_for_html(monkeypatch, tmp_path):CODE
LOWtests/test_bug_fixes.py93def test_kiwix_kill_proc_posix_signals_the_process_group():CODE
LOWtests/test_bug_fixes.py119def test_kiwix_kill_proc_without_killpg_falls_back_to_terminate():CODE
LOWtests/test_bug_fixes.py133def test_kiwix_kill_proc_skips_an_already_exited_process():CODE
LOWtests/test_bug_fixes.py146def test_chunk_article_converts_jpeg_to_png(tmp_path):CODE
LOW⚡tests/test_incremental_rerun.py23def test_missing_dir_needs_render(tmp_path):CODE
LOW⚡tests/test_incremental_rerun.py28def test_matching_source_is_reused(tmp_path):CODE
LOW⚡tests/test_incremental_rerun.py34def test_legacy_url_field_is_reused(tmp_path):CODE
LOW⚡tests/test_incremental_rerun.py41def test_shifted_position_forces_rerender_and_removes_stale_dir(tmp_path):CODE
LOW⚡tests/test_incremental_rerun.py49def test_corrupt_manifest_forces_rerender(tmp_path):CODE
LOW⚡tests/test_incremental_rerun.py58def test_corrupt_manifest_skips_article_in_chunk_stage(tmp_path):CODE
LOW⚡tests/test_article_id.py36def test_chunk_propagates_article_id_to_chunks_json(tmp_path):CODE
LOW⚡tests/test_article_id.py43def test_chunk_without_article_id_omits_it(tmp_path):CODE
LOW⚡tests/test_article_id.py52def test_gpu_scan_reads_propagated_article_id(tmp_path):CODE
LOW⚡tests/test_article_id.py60def test_gpu_scan_falls_back_to_tiles_json(tmp_path):CODE
LOW⚡tests/test_article_id.py72def test_gpu_scan_falls_back_to_numeric_dir_name(tmp_path):CODE
LOW⚡tests/test_article_id.py82def test_cpu_scan_reads_article_id_from_manifest(tmp_path):CODE
LOW⚡tests/test_article_id.py89def test_cpu_non_numeric_fallback_is_reproducible(tmp_path):CODE
LOWtests/test_local_source_text.py22def test_local_source_discovers_text_files(text_dir):CODE
LOWtests/test_local_source_text.py34def test_text_files_have_correct_metadata(text_dir):CODE
LOWtests/test_local_source_text.py56def test_source_len_includes_text_files(text_dir):CODE
LOWtests/test_atlascloud_client.py24def test_atlascloud_preserves_model_and_isolates_credentials(monkeypatch, model):CODE
LOWtests/test_atlascloud_client.py38def test_atlascloud_requires_own_key(monkeypatch, key):CODE
LOWtests/test_atlascloud_client.py57def test_atlascloud_client_routes_once(monkeypatch, error):CODE
LOWtests/test_atlascloud_client.py104def test_other_readers_keep_existing_retry_behavior(monkeypatch):CODE
LOWtests/test_extract_text.py8def test_extract_text_produces_text_md(tmp_path):CODE
LOWtests/test_extract_text.py33def test_no_extract_text_by_default(tmp_path):CODE
LOWtests/test_extract_text.py45def test_extract_text_wins_over_the_turbo_path(tmp_path):CODE
LOWtests/test_cheaperinference_client.py23def test_cheaperinference_preserves_model_and_isolates_credentials(monkeypatch, model):CODE
LOWtests/test_cheaperinference_client.py37def test_cheaperinference_requires_own_key(monkeypatch, key):CODE
LOWtests/test_cheaperinference_client.py48def test_cheaperinference_client_forwards_image_messages(monkeypatch, model):CODE
LOWtests/test_local_image_render.py45def test_local_images_produce_tile_directories(tmp_path, image_dir):CODE
LOWtests/test_local_image_render.py94def test_large_image_is_resized(tmp_path):CODE
LOWtests/test_local_image_render.py112def test_small_image_not_resized(tmp_path):CODE
LOWtests/test_pixel_query_paths.py43def test_safe_stem_has_no_separators(bad_id):CODE
LOWtests/test_pixel_query_paths.py53def test_constructed_path_stays_in_output_dir(tmp_path, bad_id):CODE
LOWtests/test_pixel_query_paths.py64def test_safe_stem_preserves_normal_ids():CODE
LOWtests/test_pixel_query_paths.py81def test_render_malicious_id_does_not_escape(tmp_path):CODE
LOWtests/test_serve_backends.py130def test_raw_search_hit_shape_and_ranking(backend):CODE
LOWtests/test_serve_backends.py149def test_article_ids_prefilter_restricts_results(backend):CODE
LOWtests/test_serve_backends.py160def test_prefilter_guarantees_k_within_department(backend):CODE
LOWtests/test_serve_backends.py168def test_filter_cache_key_reuses_positions(faiss_backend):CODE
LOW⚡tests/test_serve_backends.py181def test_reconstruct_round_trips(backend):CODE
LOW⚡tests/test_serve_backends.py188def test_reconstruct_unknown_id_is_none(backend):CODE
LOW⚡tests/test_serve_backends.py194def test_reconstruct_non_integer_id_is_none(faiss_backend):CODE
119 more matches not shown…
Unused Imports136 hits · 116 pts
SeverityFileLineSnippetContext
LOWembed/src/pixelrag_embed/embed.py55CODE
LOWembed/src/pixelrag_embed/embed.py56CODE
LOWtests/test_cli.py113CODE
LOWtests/test_cli.py114CODE
LOWtests/test_cli.py114CODE
LOWtests/test_cli.py116CODE
LOWrender/src/pixelrag_render/__init__.py6CODE
LOWrender/src/pixelrag_render/__init__.py6CODE
LOWrender/src/pixelrag_render/__init__.py6CODE
LOWrender/src/pixelrag_render/bench/bench_throughput.py20CODE
LOWrender/src/pixelrag_render/bench/__init__.py11CODE
LOWrender/src/pixelrag_render/bench/__init__.py14CODE
LOWrender/src/pixelrag_render/bench/__init__.py17CODE
LOWrender/src/pixelrag_render/bench/__init__.py20CODE
LOWrender/src/pixelrag_render/strategies/cdp_oneshot.py19CODE
LOWrender/src/pixelrag_render/strategies/cdp_fullpage.py8CODE
LOWrender/src/pixelrag_render/strategies/cdp_dc_single.py8CODE
LOW…r/src/pixelrag_render/strategies/cdp_pipelined_tabs.py12CODE
LOWrender/src/pixelrag_render/strategies/cdp_noscroll.py7CODE
LOWrender/src/pixelrag_render/strategies/cdp_sequential.py7CODE
LOW…/src/pixelrag_render/strategies/cdp_pertile_imgwait.py11CODE
LOWrender/src/pixelrag_render/strategies/__init__.py3CODE
LOWrender/src/pixelrag_render/strategies/__init__.py3CODE
LOWrender/src/pixelrag_render/strategies/__init__.py3CODE
LOWrender/src/pixelrag_render/strategies/__init__.py3CODE
LOWrender/src/pixelrag_render/strategies/__init__.py4CODE
LOWrender/src/pixelrag_render/strategies/__init__.py5CODE
LOWrender/src/pixelrag_render/strategies/__init__.py6CODE
LOWrender/src/pixelrag_render/strategies/__init__.py7CODE
LOWrender/src/pixelrag_render/strategies/__init__.py8CODE
LOWrender/src/pixelrag_render/strategies/__init__.py8CODE
LOWrender/src/pixelrag_render/strategies/__init__.py8CODE
LOWrender/src/pixelrag_render/strategies/__init__.py8CODE
LOWrender/src/pixelrag_render/strategies/cdp_directclip.py8CODE
LOWrender/src/pixelrag_render/strategies/connection.py3CODE
LOWrender/src/pixelrag_render/strategies/cdp_phased.py10CODE
LOW…der/src/pixelrag_render/strategies/cdp_pipelined_dc.py12CODE
LOWrender/src/pixelrag_render/strategies/cdp_dynamic.py7CODE
LOWrender/src/pixelrag_render/strategies/cdp_parallel.py8CODE
LOWrender/src/pixelrag_render/strategies/cdp_overlap.py12CODE
LOWrender/src/pixelrag_render/strategies/base.py8CODE
LOWrender/src/pixelrag_render/strategies/cdp_multitab.py12CODE
LOWrender/src/pixelrag_render/backends/fast_cdp.py22CODE
LOWtrain/upload_hf_dataset.py4CODE
LOWtrain/split_first5_chunks.py4CODE
LOWtrain/export_natural_filtered_v2_with_answer.py4CODE
LOWtrain/mine_text_hard_negatives.py14CODE
LOWtrain/filter_text_hard_negatives_llm.py25CODE
LOWtrain/validate_images.py11CODE
LOWtrain/run_filter_hard_negatives_chunks.py7CODE
LOWtrain/build_test_miniv8.py13CODE
LOWtrain/export_natural_filtered_v2.py4CODE
LOWtrain/run_filter_text_hard_negatives_chunks.py7CODE
LOWtrain/prepare_hf_dataset.py21CODE
LOWtrain/prepare_hf_dataset_single_jsonl.py15CODE
LOWtrain/extract_hf_image_shards.py4CODE
LOWtrain/clean_queries_simpleqa_style.py17CODE
LOWtrain/filter_hard_negatives_vqa.py22CODE
LOWtrain/upload_test_miniv7.py4CODE
LOWtrain/package_hf_image_shards.py4CODE
76 more matches not shown…
Cross-Language Confusion23 hits · 115 pts
SeverityFileLineSnippetContext
HIGHrender/src/pixelrag_render/strategies/cdp_oneshot.py45 .filter(i => !i.complete && i.loading !== 'lazy')CODE
HIGHrender/src/pixelrag_render/strategies/cdp_oneshot.py322 return r.bottom > 0 && r.top < window.innerHeight;CODE
HIGHrender/src/pixelrag_render/strategies/cdp_dc_single.py25 .filter(i => !i.complete && i.loading !== 'lazy')CODE
HIGHrender/src/pixelrag_render/strategies/cdp_dc_single.py164 return r.bottom > 0 && r.top < window.innerHeight;CODE
HIGHrender/src/pixelrag_render/strategies/cdp_sequential.py137 .filter(i => !i.complete && i.loading !== 'lazy')CODE
HIGHrender/src/pixelrag_render/strategies/cdp_sequential.py194 return r.bottom > 0 && r.top < window.innerHeight;CODE
HIGH…/src/pixelrag_render/strategies/cdp_pertile_imgwait.py168 return r.bottom > 0 && r.top < window.innerHeight;CODE
HIGHrender/src/pixelrag_render/strategies/cdp_phased.py27 .filter(i => !i.complete && i.loading !== 'lazy')CODE
HIGHrender/src/pixelrag_render/strategies/cdp_phased.py266 return r.bottom > 0 && r.top < window.innerHeight;CODE
HIGH…der/src/pixelrag_render/strategies/cdp_pipelined_dc.py32 .filter(i => !i.complete && i.loading !== 'lazy')CODE
HIGH…der/src/pixelrag_render/strategies/cdp_pipelined_dc.py280 return r.bottom > 0 && r.top < window.innerHeight;CODE
HIGHrender/src/pixelrag_render/strategies/cdp_overlap.py33 .filter(i => !i.complete && i.loading !== 'lazy')CODE
HIGHrender/src/pixelrag_render/strategies/cdp_overlap.py286 return r.bottom > 0 && r.top < window.innerHeight;CODE
HIGHrender/src/pixelrag_render/strategies/cdp_multitab.py33 .filter(i => !i.complete && i.loading !== 'lazy')CODE
HIGHrender/src/pixelrag_render/strategies/cdp_multitab.py289 return r.bottom > 0 && r.top < window.innerHeight;CODE
HIGHrender/src/pixelrag_render/backends/cdp.py392 return r.bottom > 0 && r.top < window.innerHeight;CODE
HIGHrender/src/pixelrag_render/backends/page_metrics.py33 const offset = window.scrollY || window.pageYOffset || 0;CODE
HIGHrender/src/pixelrag_render/backends/page_metrics.py40 if (r.width > 0 || r.height > 0) {CODE
HIGHrender/src/pixelrag_render/backends/fast_cdp.py71 .filter(i => !i.complete && i.loading !== 'lazy')CODE
HIGHrender/src/pixelrag_render/backends/fast_cdp.py524 return r.bottom > 0 && r.top < window.innerHeight;CODE
HIGHeval/lib/screenshot.py203 var dataSrc = img.getAttribute('data-src') || (img.dataset && img.dataset.src);CODE
HIGHeval/lib/screenshot.py204 var dataSrcset = img.getAttribute('data-srcset') || (img.dataset && img.dataset.srcset);CODE
HIGHeval/lib/screenshot.py219 .filter(function(img) { return !(img.complete && img.naturalWidth > 0); }).length;CODE
Deep Nesting143 hits · 112 pts
SeverityFileLineSnippetContext
LOWembed/src/pixelrag_embed/embed.py220CODE
LOWembed/src/pixelrag_embed/embed.py307CODE
LOWembed/src/pixelrag_embed/embed.py889CODE
LOWembed/src/pixelrag_embed/embed.py1240CODE
LOWembed/src/pixelrag_embed/embed.py1827CODE
LOWembed/src/pixelrag_embed/embed.py2240CODE
LOWembed/src/pixelrag_embed/embed.py917CODE
LOWembed/src/pixelrag_embed/embed.py1419CODE
LOWembed/src/pixelrag_embed/index.py431CODE
LOWembed/src/pixelrag_embed/chunk.py53CODE
LOWembed/src/pixelrag_embed/chunk.py67CODE
LOWembed/src/pixelrag_embed/chunk.py261CODE
LOWembed/src/pixelrag_embed/chunk.py367CODE
LOWembed/src/pixelrag_embed/embed_cpu.py62CODE
LOWrender/src/pixelrag_render/render.py192CODE
LOWrender/src/pixelrag_render/chrome.py49CODE
LOWrender/src/pixelrag_render/chrome.py170CODE
LOWrender/src/pixelrag_render/chrome.py261CODE
LOWrender/src/pixelrag_render/bench/bench_throughput.py204CODE
LOWrender/src/pixelrag_render/bench/bench_throughput.py272CODE
LOW…r/src/pixelrag_render/strategies/cdp_pipelined_tabs.py72CODE
LOW…r/src/pixelrag_render/strategies/cdp_pipelined_tabs.py174CODE
LOWrender/src/pixelrag_render/strategies/cdp_directclip.py121CODE
LOWrender/src/pixelrag_render/strategies/connection.py58CODE
LOWrender/src/pixelrag_render/strategies/cdp_phased.py151CODE
LOWrender/src/pixelrag_render/strategies/cdp_overlap.py182CODE
LOWrender/src/pixelrag_render/strategies/cdp_multitab.py85CODE
LOWrender/src/pixelrag_render/strategies/cdp_multitab.py185CODE
LOWrender/src/pixelrag_render/backends/cdp.py264CODE
LOWrender/src/pixelrag_render/backends/cdp.py405CODE
LOWrender/src/pixelrag_render/backends/fast_cdp.py315CODE
LOWrender/src/pixelrag_render/backends/fast_cdp.py232CODE
LOWrender/src/pixelrag_render/backends/fast_cdp.py435CODE
LOWtrain/export_natural_filtered_v2_with_answer.py50CODE
LOWtrain/generate_query_pairs.py363CODE
LOWtrain/mine_hard_negatives.py54CODE
LOWtrain/mine_text_hard_negatives.py69CODE
LOWtrain/filter_text_hard_negatives_llm.py223CODE
LOWtrain/filter_text_hard_negatives_llm.py540CODE
LOWtrain/validate_images.py56CODE
LOWtrain/fake_data.py32CODE
LOWtrain/export_natural_filtered_v2.py34CODE
LOWtrain/filter_hard.py78CODE
LOWtrain/filter_hard.py123CODE
LOWtrain/filter_passage.py62CODE
LOWtrain/filter_v2.py111CODE
LOWtrain/filter_v2.py151CODE
LOWtrain/clean_queries_simpleqa_style.py446CODE
LOWtrain/train_contrastors.py289CODE
LOWtrain/train_contrastors.py431CODE
LOWtrain/train_contrastors.py1029CODE
LOWtrain/train_contrastors.py1137CODE
LOWtrain/train_contrastors.py1207CODE
LOWtrain/train_contrastors.py1479CODE
LOWtrain/train_contrastors.py1762CODE
LOWtrain/train_contrastors.py561CODE
LOWtrain/train_contrastors.py698CODE
LOWtrain/train_contrastors.py1161CODE
LOWtrain/train_contrastors.py2780CODE
LOWtrain/train_contrastors.py3064CODE
83 more matches not shown…
Cross-File Repetition17 hits · 85 pts
SeverityFileLineSnippetContext
HIGHrender/src/pixelrag_render/strategies/cdp_oneshot.py0const waiteagerimgs = promise.all( array.from(document.images) .filter(i => !i.complete && i.loading !== 'lazy') .map(i STRING
HIGHrender/src/pixelrag_render/strategies/cdp_dc_single.py0const waiteagerimgs = promise.all( array.from(document.images) .filter(i => !i.complete && i.loading !== 'lazy') .map(i STRING
HIGHrender/src/pixelrag_render/strategies/cdp_sequential.py0const waiteagerimgs = promise.all( array.from(document.images) .filter(i => !i.complete && i.loading !== 'lazy') .map(i STRING
HIGHrender/src/pixelrag_render/strategies/cdp_phased.py0const waiteagerimgs = promise.all( array.from(document.images) .filter(i => !i.complete && i.loading !== 'lazy') .map(i STRING
HIGH…der/src/pixelrag_render/strategies/cdp_pipelined_dc.py0const waiteagerimgs = promise.all( array.from(document.images) .filter(i => !i.complete && i.loading !== 'lazy') .map(i STRING
HIGHrender/src/pixelrag_render/strategies/cdp_overlap.py0const waiteagerimgs = promise.all( array.from(document.images) .filter(i => !i.complete && i.loading !== 'lazy') .map(i STRING
HIGHrender/src/pixelrag_render/strategies/cdp_multitab.py0const waiteagerimgs = promise.all( array.from(document.images) .filter(i => !i.complete && i.loading !== 'lazy') .map(i STRING
HIGHrender/src/pixelrag_render/backends/fast_cdp.py0const waiteagerimgs = promise.all( array.from(document.images) .filter(i => !i.complete && i.loading !== 'lazy') .map(i STRING
HIGHrender/src/pixelrag_render/strategies/cdp_oneshot.py0new promise(resolve => {{ window.scrollto(0, {t * th}); requestanimationframe(() => requestanimationframe(() => {{ constSTRING
HIGHrender/src/pixelrag_render/strategies/cdp_dc_single.py0new promise(resolve => {{ window.scrollto(0, {t * th}); requestanimationframe(() => requestanimationframe(() => {{ constSTRING
HIGHrender/src/pixelrag_render/strategies/cdp_sequential.py0new promise(resolve => {{ window.scrollto(0, {t * th}); requestanimationframe(() => requestanimationframe(() => {{ constSTRING
HIGH…/src/pixelrag_render/strategies/cdp_pertile_imgwait.py0new promise(resolve => {{ window.scrollto(0, {t * th}); requestanimationframe(() => requestanimationframe(() => {{ constSTRING
HIGHrender/src/pixelrag_render/strategies/cdp_phased.py0new promise(resolve => {{ window.scrollto(0, {t * th}); requestanimationframe(() => requestanimationframe(() => {{ constSTRING
HIGHrender/src/pixelrag_render/strategies/cdp_overlap.py0new promise(resolve => {{ window.scrollto(0, {t * th}); requestanimationframe(() => requestanimationframe(() => {{ constSTRING
HIGHrender/src/pixelrag_render/backends/fast_cdp.py0new promise(resolve => {{ window.scrollto(0, {t * th}); requestanimationframe(() => requestanimationframe(() => {{ constSTRING
HIGH…der/src/pixelrag_render/strategies/cdp_pipelined_dc.py0new promise(resolve => {{ window.scrollto(0, {t * th}); requestanimationframe(() => requestanimationframe(() => {{ constSTRING
HIGHrender/src/pixelrag_render/strategies/cdp_multitab.py0new promise(resolve => {{ window.scrollto(0, {t * th}); requestanimationframe(() => requestanimationframe(() => {{ constSTRING
Docstring Block Structure14 hits · 70 pts
SeverityFileLineSnippetContext
HIGHrender/src/pixelrag_render/backends/pdf.py25Render a PDF to JPEG tiles. Each page is written as ``{stem}.png.tiles/tile_NNNN.jpg`` with a ``tiles.json`` maSTRING
HIGHeval/lib/simpleqa_filter.py48Load SimpleQA examples that have Wikipedia URLs. Args: num_examples: Number of Wikipedia examples to returnSTRING
HIGHeval/lib/simpleqa_filter.py91Load SimpleQA examples filtered by URL domain. Args: domain: Domain to filter by (e.g., 'wikipedia', 'arxivSTRING
HIGHeval/lib/model_config.py189 Generate output filename with model name and task included. Args: output_dir: Base output directory (eSTRING
HIGHeval/lib/retrievers.py42Build a retriever from CLI args. Args: args: Parsed argparse namespace. examples: Loaded dataset exSTRING
HIGHeval/lib/pixel_query.py144Batch-render pixel queries for a list of examples. Args: examples: List of dicts with at least ``idSTRING
HIGHeval/lib/simpleqa_data.py31Load SimpleQA dataset. Args: num_examples: Optional limit on number of examples to load. Returns: STRING
HIGHeval/lib/simpleqa_data.py72Load SimpleQA Verified dataset from Hugging Face. Args: num_examples: Optional limit on number of examples STRING
HIGHeval/lib/simpleqa_data.py797Load Natural Questions (full) split. For validation, follows the short-answer protocol used by our NQ eval: keeSTRING
HIGHeval/lib/simpleqa_data.py891Load TriviaQA rc.wikipedia validation split. Uses entity_pages.title to construct ground truth Wikipedia URLs. STRING
HIGHeval/lib/simpleqa_data.py983Load NQ-Tables dev split (table subset of Natural Questions). NQ-Tables filters Natural Questions to only keep examSTRING
HIGHeval/lib/benchmarks.py100 Load Encyclopedic VQA dataset. Args: split: Dataset split ('val' or 'test') num_examples: LimiSTRING
HIGHeval/lib/benchmarks.py210 Load Short-form QA dataset data. Args: dataset_repo: HuggingFace dataset repository name num_eSTRING
HIGHeval/lib/benchmarks.py252 Load WorldVQA dataset from HuggingFace. Args: num_examples: Limit to first N examples (optional) STRING
Structural Annotation Overuse22 hits · 34 pts
SeverityFileLineSnippetContext
LOWrender/src/pixelrag_render/strategies/cdp_fullpage.py100 # Step 1: small viewport, navigate, measure true heightCOMMENT
LOW⚡render/src/pixelrag_render/strategies/cdp_fullpage.py142 # Step 2: lock viewport-relative units before resizeCOMMENT
LOWrender/src/pixelrag_render/strategies/cdp_fullpage.py154 # Step 3: resize viewport to content height for captureCOMMENT
LOWrender/src/pixelrag_render/strategies/cdp_fullpage.py174 # Step 4: single screenshotCOMMENT
LOW⚡render/src/pixelrag_render/strategies/cdp_fullpage.py201 # Step 5: crop into tiles (NOT timed)COMMENT
LOW⚡train/train_contrastors.py312 # Step 1: get all embeddings without grad (no DDP involvement)COMMENT
LOW⚡train/train_contrastors.py318 # Step 2: compute loss, get gradient w.r.t. embeddingsCOMMENT
LOWtrain/train_contrastors.py339 # Step 3: accumulate real gradients via surrogate lossCOMMENT
LOWtrain/train_contrastors.py447 # Step 1: query embeddings come from the trainable tower; document embeddingsCOMMENT
LOWtrain/train_contrastors.py462 # Step 2: compute loss and cache only the query-side gradients.COMMENT
LOWtrain/train_contrastors.py481 # Step 3: replay only query chunks; doc tower is frozen and never backprops.COMMENT
LOWtrain/docs/synthetic_data_pipeline.md101### Step 1: Generate Visual Query PairsCOMMENT
LOWtrain/docs/synthetic_data_pipeline.md156### Step 2: Filter Self-Contained QueriesCOMMENT
LOWtrain/docs/synthetic_data_pipeline.md184### Step 3: Mine Hard NegativesCOMMENT
LOWtrain/docs/synthetic_data_pipeline.md215### Step 4: Filter Hard Negatives (VQA-based)COMMENT
LOWtrain/docs/synthetic_data_pipeline.md237### Step 5: Clean Queries (Naturalness Scoring)COMMENT
LOWtrain/docs/synthetic_data_pipeline.md256### Step 6: Export Strict Retained RowsCOMMENT
LOWtrain/docs/synthetic_data_pipeline.md273### Step 7: Split into Train/Eval/TestCOMMENT
LOWtrain/docs/synthetic_data_pipeline.md292### Step 8: Package and Upload to Hugging FaceCOMMENT
LOWeval/lib/retrieval.py2111 # Step 1: Query rewriting (if enabled)COMMENT
LOWeval/lib/retrieval.py2130 # Step 2: Build query listCOMMENT
LOWeval/lib/retrieval.py2296 # Step 3: Rerank (if reranker provided)COMMENT
Modern Structural Boilerplate30 hits · 30 pts
SeverityFileLineSnippetContext
LOWembed/src/pixelrag_embed/embed.py460def set_instruction(instruction: str) -> None:CODE
LOWrender/src/pixelrag_render/__init__.py8__all__ = ["render_file", "render_pdf", "render_url"]CODE
LOWrender/src/pixelrag_render/strategies/__init__.py15__all__ = [CODE
LOWtrain/mine_hard_negatives.py34logger = logging.getLogger(__name__)CODE
LOWtrain/mine_text_hard_negatives.py31logger = logging.getLogger(__name__)CODE
LOWtrain/fake_data.py23logger = logging.getLogger(__name__)CODE
LOWtrain/filter_hard.py32logger = logging.getLogger(__name__)CODE
LOWtrain/filter_entity_queries.py30logger = logging.getLogger(__name__)CODE
LOWtrain/filter_passage.py28logger = logging.getLogger(__name__)CODE
LOWtrain/filter_v2.py23logger = logging.getLogger(__name__)CODE
LOWtrain/verify_embeddings.py28logger = logging.getLogger(__name__)CODE
LOW⚡train/train_contrastors.py64logger = logging.getLogger(__name__)CODE
LOWtrain/filter_strict.py27logger = logging.getLogger(__name__)CODE
LOWtrain/dataset.py10logger = logging.getLogger(__name__)CODE
LOWtrain/train.py54logger = logging.getLogger(__name__)CODE
LOWtrain/evaluate.py15logger = logging.getLogger(__name__)CODE
LOWindex/src/pixelrag_index/distributed.py26logger = logging.getLogger(__name__)CODE
LOWindex/src/pixelrag_index/sources/kiwix.py19logger = logging.getLogger(__name__)CODE
LOWeval/run_bench.py38logger = logging.getLogger(__name__)CODE
LOWeval/lib/simpleqa_filter.py9logger = logging.getLogger(__name__)CODE
LOWeval/lib/dump_retrieval.py26logger = logging.getLogger(__name__)CODE
LOWeval/lib/retrievers.py36logger = logging.getLogger(__name__)CODE
LOWeval/lib/pixel_query.py18logger = logging.getLogger(__name__)CODE
LOWeval/lib/__init__.py64__all__ = [CODE
LOWeval/lib/llm.py36logger = logging.getLogger(__name__)CODE
LOWeval/lib/retrieval.py18logger = logging.getLogger(__name__)CODE
LOWeval/lib/screenshot.py15logger = logging.getLogger(__name__)CODE
LOW⚡eval/lib/simpleqa_data.py22logger = logging.getLogger(__name__)CODE
LOWeval/lib/benchmarks.py22logger = logging.getLogger(__name__)CODE
LOWserve/src/pixelrag_serve/backends.py27 def set_nprobe(self, n: int | None) -> None: ...CODE
Modern AI Meta-Vocabulary9 hits · 28 pts
SeverityFileLineSnippetContext
MEDIUM⚡embed/src/pixelrag_embed/embed.py1823# Multi-GPU orchestrationCOMMENT
MEDIUMtrain/generate_text_query_pairs.py118# Built-in few-shot examples (used when --fewshot-file is not provided)COMMENT
MEDIUMtrain/sft/push_multi3_to_hf.py114## Data — retrieval-augmented multi-imageCOMMENT
MEDIUMtrain/sft/push_multik_to_hf.py141## Data — variable-k retrieval-augmented multi-imageCOMMENT
MEDIUMtrain/sft/push_3x_v5_snapshot.py127## Data — variable-k retrieval-augmented multi-image (with 2× top3 weighting)COMMENT
MEDIUMeval/run_bench.py1485 # Text RAG specificCOMMENT
MEDIUMeval/lib/llm.py291 # Text retrieval rendered as images. Mirror the text-RAG framing soCOMMENT
MEDIUMeval/lib/llm.py348 # Few-shot as prior user/assistant turns (canonical chat few-shot format)COMMENT
MEDIUMserve/src/pixelrag_serve/api.py755 # Load embedding modelCOMMENT
Verbosity Indicators14 hits · 22 pts
SeverityFileLineSnippetContext
LOWrender/src/pixelrag_render/strategies/cdp_fullpage.py100 # Step 1: small viewport, navigate, measure true heightCOMMENT
LOW⚡render/src/pixelrag_render/strategies/cdp_fullpage.py142 # Step 2: lock viewport-relative units before resizeCOMMENT
LOWrender/src/pixelrag_render/strategies/cdp_fullpage.py154 # Step 3: resize viewport to content height for captureCOMMENT
LOWrender/src/pixelrag_render/strategies/cdp_fullpage.py174 # Step 4: single screenshotCOMMENT
LOW⚡render/src/pixelrag_render/strategies/cdp_fullpage.py201 # Step 5: crop into tiles (NOT timed)COMMENT
LOW⚡train/train_contrastors.py312 # Step 1: get all embeddings without grad (no DDP involvement)COMMENT
LOW⚡train/train_contrastors.py318 # Step 2: compute loss, get gradient w.r.t. embeddingsCOMMENT
LOWtrain/train_contrastors.py339 # Step 3: accumulate real gradients via surrogate lossCOMMENT
LOWtrain/train_contrastors.py447 # Step 1: query embeddings come from the trainable tower; document embeddingsCOMMENT
LOWtrain/train_contrastors.py462 # Step 2: compute loss and cache only the query-side gradients.COMMENT
LOWtrain/train_contrastors.py481 # Step 3: replay only query chunks; doc tower is frozen and never backprops.COMMENT
LOWeval/lib/retrieval.py2111 # Step 1: Query rewriting (if enabled)COMMENT
LOWeval/lib/retrieval.py2130 # Step 2: Build query listCOMMENT
LOWeval/lib/retrieval.py2296 # Step 3: Rerank (if reranker provided)COMMENT
Redundant / Tautological Comments14 hits · 18 pts
SeverityFileLineSnippetContext
LOWrender/src/pixelrag_render/strategies/cdp_dynamic.py149 # Check if scroll movedCOMMENT
LOWtrain/train_swift.py216 # Print resultsCOMMENT
LOWindex/src/pixelrag_index/distributed.py103 # Check if already claimedCOMMENT
LOWeval/run_livevqa.py617 # Check if we have any retrieved tiles (beyond just the editorial photo)COMMENT
LOWeval/run_bench.py817 # Check if any retrieved tile's URL matches the ground truth URLCOMMENT
LOWeval/run_bench.py820 # Check if the ground truth URL is contained in the retrieved URLsCOMMENT
LOWeval/run_bench.py1102 # Check if output file existsCOMMENT
LOWeval/lib/retrieval.py1033 # Check if it's a Wikipedia URLCOMMENT
LOWeval/lib/retrieval.py1741 # Check if index existsCOMMENT
LOWeval/lib/screenshot.py114 # Check if image changed (detect end of scrolling)COMMENT
LOWeval/lib/screenshot.py328 # Check if compression neededCOMMENT
LOWeval/lib/simpleqa_data.py337 # Check if valid screenshot already existsCOMMENT
LOWeval/lib/simpleqa_data.py678 # Check if tiles already exist for this imageCOMMENT
LOWeval/lib/benchmarks.py636 # Check if images directory exists for optional image loadingSTRING
Magic Placeholder Names3 hits · 15 pts
SeverityFileLineSnippetContext
HIGHeval/README.md200export ATLASCLOUD_API_KEY=your-api-keyCODE
HIGHeval/README.md258export MINIMAX_API_KEY=your-api-keyCODE
HIGHeval/lib/model_config.py159 "Set it with: export GOOGLE_API_KEY='your-api-key' or export GEMINI_API_KEY='your-api-key' and GOOGLCODE
AI Slop Vocabulary5 hits · 14 pts
SeverityFileLineSnippetContext
MEDIUM⚡render/src/pixelrag_render/bench/bench_throughput.py383# Bench: clean harness that takes any CaptureStrategyCOMMENT
MEDIUMeval/pyproject.toml6# The base harness (run_bench.py + lib/) is a pure HTTP client — retrieval and reader runCOMMENT
MEDIUMeval/lib/retrieval.py1191 # Use a more robust approach: continue even if some screenshots failCOMMENT
LOWeval/lib/retrieval.py2151 # Single or no image: just use the standard pathCOMMENT
LOWeval/lib/screenshot.py156 # Only one screenshot, just use itCOMMENT
AI Structural Patterns15 hits · 10 pts
SeverityFileLineSnippetContext
LOWembed/src/pixelrag_embed/embed.py763CODE
LOWembed/src/pixelrag_embed/embed.py1827CODE
LOWembed/src/pixelrag_embed/embed.py2364CODE
LOWembed/src/pixelrag_embed/index.py289CODE
LOWrender/src/pixelrag_render/backends/cdp.py908CODE
LOWeval/run_bench.py514CODE
LOWeval/lib/pixel_query.py184CODE
LOWeval/lib/llm.py466CODE
LOWeval/lib/retrieval.py1234CODE
LOWeval/lib/retrieval.py1376CODE
LOWeval/lib/retrieval.py1528CODE
LOWeval/lib/retrieval.py2041CODE
LOWeval/lib/retrieval.py2462CODE
LOWeval/lib/retrieval.py3057CODE
LOWeval/lib/retrieval.py3417CODE
Over-Commented Block8 hits · 8 pts
SeverityFileLineSnippetContext
LOWdeploy/api-switch.sh1#!/usr/bin/env bashCOMMENT
LOWrender/src/pixelrag_render/backends/cdp.py181 return r.get("result", {})COMMENT
LOWrender/src/pixelrag_render/backends/page_metrics.py21# is only as tall as the document when the page lets it size to its content;COMMENT
LOWtrain/run_generate_batches.sh1#!/bin/bashCOMMENT
LOWtrain/recipes/v8s_ablation.sh1#!/usr/bin/env bashCOMMENT
LOWeval/reproduce.sh1#!/bin/bashCOMMENT
LOWeval/dump_bench.sh1#!/bin/bashCOMMENT
LOWeval/serve_up.sh1#!/bin/bashCOMMENT
Example Usage Blocks5 hits · 8 pts
SeverityFileLineSnippetContext
LOWrender/chrome-build/build-headless-shell.sh6# Usage:COMMENT
LOWtrain/run_generate_batches.sh7# Usage:COMMENT
LOWtrain/recipes/v8s_ablation.sh9# Usage:COMMENT
LOWtrain/sft/train_qwen3vl_3x_scratch.yaml2### Usage:COMMENT
LOWtrain/sft/eval_fanout.sh3# Usage:COMMENT
Self-Referential Comments2 hits · 6 pts
SeverityFileLineSnippetContext
MEDIUMtests/test_bug_fixes.py151 # Create a small JPEG tileCOMMENT
MEDIUMrender/src/pixelrag_render/strategies/cdp_overlap.py109 # Create a dummy proc wrapper for tab_b (so close() doesn't double-kill)COMMENT
Fake / Example Data5 hits · 6 pts
SeverityFileLineSnippetContext
LOW⚡tests/test_render.py27 + "<p>lorem ipsum dolor sit amet. </p>" * 60CODE
LOW⚡tests/test_render.py27 + "<p>lorem ipsum dolor sit amet. </p>" * 60CODE
LOWtrain/fake_data.py8 uv run python training/fake_data.py \STRING
LOWtrain/CLAUDE.md273`fake_data.py` generates template-based queries from article titles (not LLM-augmented):CODE
LOWtrain/CLAUDE.md275uv run python fake_data.py \CODE