refactor!: delete the dead code M53 was right about, after checking all of it

docs/superpowers/follow-ups-after-4.0.0-remediation.md

@@ -112,10 +112,22 @@ Verified against current source:
 | `constants.py` `check_connectivity` hardcodes 1.25/1.1 | recheck; the surrounding code has moved |
 | `isomers/parallel_embed.py` 138 lines | claim stands, but see the decision needed above |
 
-The remaining nine entries were not re-checked and must be before anything is
-deleted. This is the third time a manifest entry has turned out to be already
-closed or wrong; the pattern is now reliable enough to treat every entry as a
-claim rather than a fact.
+**Full re-verification, 2026-08-03.** All 13 entries checked against source. The
+list was **substantially wrong** — only 4 of 13 described code that was both
+present and dead:
+
+| verdict | entries |
+|---|---|
+| **already deleted** by earlier phases | `pad_molecular_batch`, `cli/progress` `create_progress`, `IsomerProgressCallback` |
+| **not dead — M53 wrong** | `utils/stereo_check` (6 live uses), `cli/results` `FailedMolecule` + `print_failures` (live from `run.py` since the C6/C7 reconciliation work — the finding's "run.py admits failures is always []" no longer holds), `ASE/thermo` `mol2atoms`, `STANDARD_PRESSURE`, `isomers/parallel_embed` |
+| **genuinely dead — deleted** | `utils_file.py` (whole module), `count_from_output`, `encode_smiles`, `decode_smiles`, `housekeeping_helper`, and 3 constants (`BOND_STRETCH_TOLERANCE`, `COLLISION_THRESHOLD`, `SUPPORTED_MODELS`) |
+| **unresolved — line numbers stale** | `exceptions.py` "4 classes never raised" (line 41 is `OptimizationError`, raised 3x, so the cited lines no longer point at what the finding describes); `model_wrapper`'s legacy `name` API; `ASE/thermo` `model_name` param |
+
+Net: **256 lines removed from `src/`, 167 from `tests/`** — not the ~450 of `src/`
+the finding claimed, because a third of it was gone and half of the rest is alive.
+
+The three unresolved entries need their own pass; the audit's line numbers cannot
+be used to find them.
 
 ## What the remediation closed
 

src/Auto3D/cli/results.py

@@ -105,18 +105,6 @@ def print_failures(failures: list[FailedMolecule], verbose: bool = False) -> Non
         console.print("[dim]Run with -v to see details[/dim]")
 
 
-def count_from_output(output_path: str) -> tuple[int, int]:
-    """Return (unique_molecule_count, conformer_count) from an output SDF.
-
-    Thin back-compat wrapper around :func:`Auto3D.results.count_output` (the
-    single source of truth). ``main()`` now returns a ``WorkflowResult`` that
-    carries these counts, so the CLI reads them off the result instead.
-    """
-    from Auto3D.results import count_output
-
-    return count_output(output_path)
-
-
 def output_json(results: WorkflowResults) -> None:
     """Output results as JSON.
 

src/Auto3D/constants.py

@@ -7,8 +7,6 @@
 HARTREE_TO_KCAL_PER_MOL = 627.50947337481  # 1 Hartree in kcal/mol
 
 # Geometry thresholds
-BOND_STRETCH_TOLERANCE = 1.25  # Maximum bond stretch factor
-COLLISION_THRESHOLD = 1.1  # Å, minimum distance for clash detection
 MIN_ATOM_DISTANCE = 0.9  # Å, minimum allowed interatomic distance
 
 # Conformer generation limits
@@ -43,9 +41,6 @@
 MODEL_ANI2XT = "ANI2xt"
 
 # Supported model names (for validation)
-SUPPORTED_MODELS = frozenset({MODEL_AIMNET, MODEL_ANI2X, MODEL_ANI2XT})
-
-# Backward-compatible alias: "AIMNET" now maps to the aimnet registry default.
 DEFAULT_AIMNET_MODEL = "aimnet2"
 # Built-in (non-aimnet) engines kept for back-compat.
 BUILTIN_ANI_MODELS = frozenset({MODEL_ANI2X.upper(), MODEL_ANI2XT.upper()})

src/Auto3D/utils/__init__.py

@@ -30,14 +30,11 @@
     combine_smi,
     create_chunk_meta_names,
     decode_ids,
-    decode_smiles,
     encode_ids,
-    encode_smiles,
     guess_file_type,
     hash_enumerated_smi_IDs,
     hash_taut_smi,
     housekeeping,
-    housekeeping_helper,
     reorder_sdf,
 )
 from Auto3D.utils.logging_config import configure_logging, get_logger
@@ -99,11 +96,8 @@
     "check_valid_configuration",
     # File operations
     "guess_file_type",
-    "encode_smiles",
-    "decode_smiles",
     "hash_enumerated_smi_IDs",
     "hash_taut_smi",
-    "housekeeping_helper",
     "housekeeping",
     "create_chunk_meta_names",
     "combine_smi",

src/Auto3D/utils/file_ops.py

@@ -12,9 +12,7 @@
 """
 from __future__ import annotations
 
-import base64
 import collections
-import hashlib
 import os
 import shutil
 from collections import defaultdict
@@ -176,135 +174,6 @@ def guess_file_type(filename: str) -> str:
     return Path(filename).suffix[1:]
 
 
-def encode_smiles(smiles: str, max_length: int = 50) -> str:
-    """Encode a SMILES string for use in filenames.
-
-    Transforms a SMILES string into a filesystem-safe string by replacing
-    special characters with alphanumeric equivalents. For SMILES longer
-    than max_length, uses a hash-based encoding.
-
-    The encoding maps common SMILES characters to filename-safe alternatives:
-    - '=' -> 'd' (double bond)
-    - '#' -> 't' (triple bond)
-    - '@' -> 'a' (stereochemistry)
-    - '/' -> 's' (cis/trans)
-    - '\\' -> 'b' (cis/trans)
-    - '+' -> 'p' (positive charge)
-    - '-' -> 'm' (negative charge)
-    - '(' -> 'L' (left paren)
-    - ')' -> 'R' (right paren)
-    - '[' -> 'K' (left bracket)
-    - ']' -> 'J' (right bracket)
-    - '%' -> 'X' (ring number indicator)
-
-    Args:
-        smiles: The SMILES string to encode.
-        max_length: Maximum length for encoded string before using hash.
-            Defaults to 50 characters.
-
-    Returns:
-        A filesystem-safe encoded string representing the SMILES.
-
-    Example:
-        >>> encode_smiles("CCO")
-        'CCO'
-        >>> encode_smiles("C=C")
-        'CdC'
-        >>> encode_smiles("C#N")
-        'CtN'
-        >>> encode_smiles("[NH4+]")
-        'KNH4pJ'
-    """
-    # Define character replacements for filesystem safety
-    # Using single lowercase letters that are unlikely to cause collisions
-    replacements = {
-        '=': 'd',   # double bond
-        '#': 't',   # triple bond
-        '@': 'a',   # stereochemistry
-        '/': 's',   # cis/trans
-        '\\': 'b',  # cis/trans (backslash)
-        '+': 'p',   # positive charge
-        '-': 'm',   # negative charge (also single bond, but rare in SMILES)
-        '(': 'L',   # left parenthesis
-        ')': 'R',   # right parenthesis
-        '[': 'K',   # left bracket
-        ']': 'J',   # right bracket
-        '%': 'X',   # ring number indicator (for rings > 9)
-    }
-
-    # Apply replacements
-    encoded = smiles
-    for char, replacement in replacements.items():
-        encoded = encoded.replace(char, replacement)
-
-    # If still too long, use a hash-based encoding
-    if len(encoded) > max_length:
-        # Use SHA256 hash truncated to produce a shorter, unique identifier
-        hash_obj = hashlib.sha256(smiles.encode('utf-8'))
-        # Take first 16 characters of base64-encoded hash (url-safe)
-        hash_str = base64.urlsafe_b64encode(hash_obj.digest()[:12]).decode('utf-8')
-        # Combine a prefix of the encoded SMILES with the hash
-        prefix_len = max_length - len(hash_str) - 1  # -1 for separator
-        if prefix_len > 0:
-            encoded = f"{encoded[:prefix_len]}_{hash_str}"
-        else:
-            encoded = hash_str
-
-    return encoded
-
-
-def decode_smiles(encoded: str) -> str:
-    """Decode an encoded SMILES string back to the original SMILES.
-
-    Reverses the encoding performed by encode_smiles for short SMILES strings.
-    Note: For hash-encoded (long) SMILES, the original cannot be recovered.
-
-    Args:
-        encoded: The encoded SMILES string.
-
-    Returns:
-        The decoded SMILES string. For hash-encoded strings, returns the
-        input unchanged since the original cannot be recovered.
-
-    Example:
-        >>> decode_smiles("CdC")
-        'C=C'
-        >>> decode_smiles("CtN")
-        'C#N'
-        >>> decode_smiles("KNH4pJ")
-        '[NH4+]'
-    """
-    # Define reverse replacements
-    # Order matters: longer replacements should not interfere with shorter ones
-    replacements = {
-        'd': '=',   # double bond
-        't': '#',   # triple bond
-        'a': '@',   # stereochemistry
-        's': '/',   # cis/trans
-        'b': '\\',  # cis/trans (backslash)
-        'p': '+',   # positive charge
-        'm': '-',   # negative charge
-        'L': '(',   # left parenthesis
-        'R': ')',   # right parenthesis
-        'K': '[',   # left bracket
-        'J': ']',   # right bracket
-        'X': '%',   # ring number indicator
-    }
-
-    # Check if this looks like a hash-encoded string (contains underscore near end
-    # followed by base64-like characters)
-    if '_' in encoded and len(encoded.split('_')[-1]) >= 12:
-        # Likely hash-encoded, can't decode
-        return encoded
-
-    # Apply reverse replacements
-    decoded = encoded
-    for char, replacement in replacements.items():
-        decoded = decoded.replace(char, replacement)
-
-    return decoded
-
-
 def hash_enumerated_smi_IDs(smi: str, out: str) -> None:
     """Write all SMILES with hashed IDs into a new file.
 
@@ -368,23 +237,6 @@ def hash_taut_smi(smi: str, out: str) -> None:
             f.write(molecule)
 
 
-def housekeeping_helper(folder: str, file: str) -> None:
-    """Move a file into the specified folder.
-
-    Args:
-        folder: Destination folder path.
-        file: Path to the file to move.
-
-    Returns:
-        None. Moves the file to the destination folder.
-
-    Example:
-        >>> housekeeping_helper("/tmp/output", "/tmp/results.sdf")
-    """
-    new_name = Path(folder) / Path(file).name
-    shutil.move(file, str(new_name))
-
-
 def housekeeping(job_name: str, folder: str, optimized_structures: str) -> None:
     """Move this job directory's metadata files into a folder.
 

tests/test_cli_results.py

@@ -96,19 +96,3 @@ def test_output_json():
     output_json(results)
 
 
-def test_count_from_output_counts_molecules_and_conformers(tmp_path):
-    from rdkit import Chem
-    from rdkit.Chem import AllChem
-    from Auto3D.cli.results import count_from_output
-
-    out = tmp_path / "out.sdf"
-    with Chem.SDWriter(str(out)) as w:
-        for name in ["a", "a", "b"]:  # 2 unique ids, 3 conformers
-            m = Chem.AddHs(Chem.MolFromSmiles("CCO"))
-            AllChem.EmbedMolecule(m, randomSeed=1)
-            m.SetProp("_Name", name)
-            w.write(m)
-
-    molecules, conformers = count_from_output(str(out))
-    assert molecules == 2
-    assert conformers == 3