allaun/Research-Stack
1
0
Fork 0
mirror of https://github.com/allaunthefox/Research-Stack.git synced 2026-07-31 03:05:21 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
Research-Stack/audit/consistency_audit.md
Allaun Silverfox 8e51acad08 audit: full codebase inspection — 6 auditors, 9,606 objects, 2,068 lines of findings 
CRITICAL (7):
- C1: Hardcoded Wolfram Alpha API key (revoke immediately)
- C2: Q16_16 rounding diverges Lean↔Python↔C (data corruption)
- C3: Receipts describe non-existent files (fabrication)
- C4: '3,500+ proofs' = compilation units, not theorems
- C5: NaN sentinel collision Lean↔C (silent corruption)
- C6: eigensolid_convergence theorem is a tautology
- C7: 74% CI failure rate

HIGH (12): duplicate definitions, SQL injection, command injection,
  1,029 duplicate files, 66 large files in git, AGENTS.md drift

MEDIUM (23): stale docs, broken cross-refs, unhandled errors
LOW (31): naming violations, misplaced files, cleanup

Master synthesis: audit/MASTER_AUDIT_SYNTHESIS.md
Per-dimension reports: audit/*_audit.md
2026-06-21 02:07:35 -05:00

12 KiB
Raw Blame History

Consistency Audit: Research-Stack Repository

Repository: https://github.com/allaunthefox/Research-Stack Branch: main Audit Date: 2025-06-21 Scope: ALL 8,449 tracked files across entire repository

Summary

Category Count Severity
Naming convention violations 220 Medium
Misplaced files 111 High
Version skew issues 5 Medium
Dead symlinks 2 Low
Duplicate content sets 441 High
Removed feature references 4 Low
Gitignore gaps 179 Medium

Total issues: 962 Critical: 0 blocking High: 2 categories (misplaced files, duplicates)

Critical Issues

None identified. No blocking inconsistencies that prevent builds or cause data loss.

High Severity Issues

H1: NoDupeLabs Code Duplication Across Layers (530 + 499 files)

The NoDupeLabs project exists as near-identical copies in both 4-Infrastructure/NoDupeLabs/ (530 files) and 5-Applications/nodupe/ (499 files). This is the single largest source of repository bloat from iteration drift.

Evidence (git blob hashes match):

  • 4-Infrastructure/NoDupeLabs/tests/core/test_database_comprehensive.py == 5-Applications/nodupe/tests/core/test_database_comprehensive.py (109KB each)
  • 4-Infrastructure/NoDupeLabs/.integrity/SHA512SUMS.txt == 5-Applications/nodupe/.integrity/SHA512SUMS.txt (101KB each)
  • 4-Infrastructure/NoDupeLabs/tests/core/test_coverage_100_percent.py == 5-Applications/nodupe/tests/core/test_coverage_100_percent.py (84KB each)
  • 100+ more files with identical SHA hashes

Recommended action: Determine which is canonical (likely 4-Infrastructure/NoDupeLabs/ per ARCHITECTURE.md) and remove the 5-Applications/nodupe/ copy. Add a README explaining the canonical location.

H2: 26MB GIF Duplicated in text-to-cad

5-Applications/text-to-cad/skills/cad/assets/text-to-cad-demo.gif (25.1MB) and 5-Applications/text-to-cad/skills/urdf/assets/text-to-cad-demo.gif (25.1MB) are identical content. Total waste: 26MB.

Recommended action: Symlink or reference a single canonical asset location.

H3: 111 Misplaced Files Across Layers

Files are located in directories that do not match their layer's architectural purpose per ARCHITECTURE.md. The largest category is documentation markdown files (99 files) misplaced in 0-Core-Formalism/ (24 files) and 5-Applications/ (75 files).

Medium Severity Issues

M1: Naming Convention Violations (220 files)

AGENTS.md convention: PascalCase types, camelCase functions. Lean modules should use PascalCase.

Lean files with lowercase_snake names (213 files):

  • Test/debug files: debug_eval.lean, test_native_decide.lean, braided_field_sim.lean
  • AlphaProofNexus adapters: 100+ files like erdos_12.parts.i.lean, erdos_125.variants.positive_lower_density.lean, graph_conjecture2.lean, green_57.lean
  • Extension scaffold files: bipartite_reconstruction.lean, green_tao_5_4.lean, etc.

Python files with PascalCase names (7 files):

  • 1-Distributed-Systems/agents/claw/src/QueryEngine.py
  • 1-Distributed-Systems/agents/claw/src/Tool.py
  • 2-Search-Space/search/whoogle-search/misc/update-translations.py
  • 2-Search-Space/simulations/Newtonian-Superfluid-Simulation/AtomSim.py
  • 2-Search-Space/simulations/Newtonian-Superfluid-Simulation/Galaxy.py
  • 2-Search-Space/simulations/Newtonian-Superfluid-Simulation/GalaxyRing.py
  • 2-Search-Space/simulations/Newtonian-Superfluid-Simulation/Universe.py

Recommended action: Rename Lean modules to PascalCase. Rename Python modules to snake_case per PEP 8.

M2: Gitignore Gaps (179 files)

The .gitignore file is comprehensive (260+ patterns) but 179 files that match gitignore patterns are still committed:

Category Count Examples
AI tool configs 11 .vscode/, .cursor/, .devin/, .roo/, .windsurf/ files
shared-data/ directory 164 Receipts, benchmarks, blockchain corpus
Root symlinks 1 result-devcontainer
Quarantine files 3 .quarantine files

Recommended action: These files were likely force-added via git add -f or added before patterns were created. Run git rm --cached for truly generated files. For shared-data/, either remove from git (if truly generated) or remove from .gitignore.

M3: Version Skew - Multiple lakefile Formats (5 files)

Five lakefile.* files exist with different formats, suggesting version drift:

File Format Status
0-Core-Formalism/lean/Semantics/lakefile.toml TOML Canonical - main workspace
0-Core-Formalism/otom/tools/lean/Semantics/lakefile.lean Lean Legacy format - should migrate to TOML
0-Core-Formalism/lean/external/OTOM/lakefile.toml TOML External dependency
0-Core-Formalism/lean/singer-theorem-lean/lakefile.toml TOML Separate project
lean_binned/lakefile.toml TOML Separate project

Recommended action: Migrate lakefile.lean to lakefile.toml format for consistency.

M4: Duplicate Semantics Module Names (32 pairs)

32 Lean modules have duplicate filenames in different directories within 0-Core-Formalism/lean/Semantics/. These may have diverged or be stale copies:

Filename Locations
Bind.lean Semantics/ and Semantics/Components/
Core.lean ExtensionScaffold/Compression/ and Semantics/Components/ and Semantics/HCMMR/ and Semantics/InvariantReceipt/ and Semantics/TMMCP/
Bridge.lean Semantics/Adapters/AlphaProofNexus/ and Semantics/HCMMR/
ReceiptCore.lean Semantics/ and otom/tools/lean/Semantics/Semantics/
Constitution.lean Semantics/ and otom/tools/lean/Semantics/Semantics/
FAMM.lean Semantics/ and otom/tools/lean/Semantics/Semantics/

Recommended action: Audit each pair to determine which is canonical. Remove or rename stale copies.

Low Severity Issues

L1: Dead Symlinks (2)

Symlink Target Issue
4-Infrastructure/bin/share /home/allaun/Research Stack/tools/bin/nextpnr-himbaechel-share Absolute path to author's local machine
result-devcontainer /nix/store/iajh1kzizdnb8ih88wd0rczdqmyg456v-research-stack-otom.tar.gz Absolute nix store path (GC may delete)

Recommended action: Replace absolute symlinks with relative paths or add a setup script.

L2: Legacy Directory Content (4 files)

0-Core-Formalism/lean/Semantics/legacy/ contains 4 files that reference deprecated concepts. The .gitignore explicitly excludes 0-Core-Formalism/lean/Semantics/Ancillary/LegacyQuarantine/ but these are in a different legacy/ directory.

Files:

  • 0-Core-Formalism/lean/Semantics/legacy/6point5sigma/HamiltonianMechanics.lean
  • 0-Core-Formalism/lean/Semantics/legacy/6point5sigma/plan.md
  • 0-Core-Formalism/lean/Semantics/legacy/6point5sigma/HamiltonianMechanics.graphml
  • 0-Core-Formalism/lean/Semantics/legacy/RealityContractMassNumber.lean

L3: OpenClaw References

OpenClaw was decommissioned per AGENTS.md (2026-06-20). No references found in code. The .gitignore correctly excludes 5-Applications/tools-scripts/external/openclaw/. Clean removal verified.

L4: Large Files in Repository (67 files > 1MB)

The repository contains 67 files larger than 1MB (total: ~370MB). Git does not handle large files well. Consider using Git LFS for:

  • shared-data/papers/2026-05/2511.21459.pdf (44.9MB)
  • 5-Applications/text-to-cad/skills/*/assets/text-to-cad-demo.gif (25.1MB each)
  • 3-Mathematical-Models/equations_compressed/*.compressed (24.2MB)
  • 4-Infrastructure/hardware/cff_accelerator.json (13.8MB)
  • 11 WAV audio files in 2-Search-Space/simulations/matter-frequencies/ (11MB each)

Per-Directory Breakdown

0-Core-Formalism/ (1,602 files) - Core Formalism Layer

Issues found:

  • 24 markdown docs that should be in 6-Documentation/
  • 213 Lean files with lowercase_snake names (should be PascalCase)
  • 1 AGENTS.md that belongs at repo root
  • 4 legacy files referencing deprecated concepts
  • 32 duplicate module names (same filename in different subdirs)
  • 2 lakefiles with potential version skew
  • 1 Rust project (ftdi_gimbal/) in formalism layer
  • Quarantine files: Q064Experimental.lean, 3 .quarantine files

1-Distributed-Systems/ (207 files)

Issues found:

  • 2 Python files with PascalCase names (QueryEngine.py, Tool.py)
  • Cargo.toml workspace nesting is normal (workspace with member crates)

2-Search-Space/ (1,627 files)

Issues found:

  • 4 Python files with PascalCase names in Newtonian-Superfluid-Simulation/
  • 1 Python file with hyphens (update-translations.py)
  • Large WAV files (11MB each) should use Git LFS
  • Duplicate SVG assets (big-logo.svg / biglogo-with-text.svg)

3-Mathematical-Models/ (171 files)

Issues found:

  • 1 large compressed file (24.2MB) should use Git LFS

4-Infrastructure/ (1,419 files)

Issues found:

  • 1 dead symlink (bin/share -> absolute path)
  • 530 NoDupeLabs files duplicated in 5-Applications/nodupe/
  • 1 large JSON file (hardware/cff_accelerator.json, 13.8MB)

5-Applications/ (1,281 files)

Issues found:

  • 1 Python script at root (fix_offloat.py) should be here
  • 75 markdown docs (hutter_prize docs) that should be in 6-Documentation/
  • 499 NoDupeLabs files duplicated from 4-Infrastructure/NoDupeLabs/
  • 1 dead symlink (result-devcontainer)
  • 26MB GIF duplicated in text-to-cad
  • Verilog files (out/verilog/*.v) should be in 4-Infrastructure/hardware/

6-Documentation/ (1,586 files)

Issues found:

  • 6 Lean files in docs/geoweird/ that should be in 0-Core-Formalism/lean/
  • 2 Lean files in invention_record/ that should be in 0-Core-Formalism/lean/
  • 1 quarantine file (DRIFT_QUARANTINE_YANG_MILLS_SPEC.md)

6-Kernel-Shim/ (not in main tree count)

Issues found:

  • 2 C kernel modules (pist_compress.c, pist_neuromorphic.c) should be in 4-Infrastructure/

Root Level (96 files)

Issues found:

  • 1 Python script (fix_offloat.py) at root
  • 1 dead symlink (result-devcontainer)
  • 1 data symlink (data)
  • 2 duplicate README.md entries at root (git artifact)

Dot-Directories (88 files)

Issues found:

  • 11 AI tool config files committed despite being in .gitignore
  • .devin/skills/lean-proof/SKILL.md is intentionally committed (skill for agents) but parent .devin/ is gitignored -- this creates confusion

Recommendations by Priority

Immediate (this week)

  1. Remove NoDupeLabs duplicate from 5-Applications/nodupe/ (499 files, ~2MB)
  2. Deduplicate text-to-cad GIF (save 26MB)
  3. Fix dead symlinks - make relative or document

Short-term (this month)

  1. Move markdown docs from 0-Core-Formalism/ and 5-Applications/ to 6-Documentation/
  2. Move C kernel modules from 6-Kernel-Shim/ to 4-Infrastructure/
  3. Clean up .quarantine files (3 files)
  4. Migrate lakefile.lean to lakefile.toml

Medium-term (next quarter)

  1. Audit 32 duplicate Semantics modules - determine canonical versions
  2. Rename Lean files to PascalCase (213 files)
  3. Rename Python files to snake_case (7 files)
  4. Set up Git LFS for large files (>1MB)
  5. Clean up shared-data/ - either remove from git or remove from .gitignore

Appendix: Methodology

  1. Read AGENTS.md for naming conventions
  2. Read ARCHITECTURE.md for layer structure
  3. Parsed audit_file_tree.json (8,449 files)
  4. Queried GitHub API for git tree blob hashes (9,661 objects)
  5. Queried GitHub API for symlink targets
  6. Checked .gitignore against committed files
  7. Cross-referenced file paths against layer architecture