dayhoff-tools 1.14.6__py3-none-any.whl → 1.14.8__py3-none-any.whl

dayhoff_tools/batch/workers/boltz.py

@@ -88,7 +88,9 @@ class BoltzProcessor:
         match = re.match(pattern1, base_name)
         if match:
             protein_id = match.group(1)
-            logger.debug(f"Extracted protein ID '{protein_id}' from '{filename}' (pattern 1)")
+            logger.debug(
+                f"Extracted protein ID '{protein_id}' from '{filename}' (pattern 1)"
+            )
             return protein_id
 
         # Pattern 2: PROTEINID_suffix (no leading number)
@@ -96,7 +98,9 @@ class BoltzProcessor:
         match = re.match(pattern2, base_name)
         if match:
             protein_id = match.group(1)
-            logger.debug(f"Extracted protein ID '{protein_id}' from '{filename}' (pattern 2)")
+            logger.debug(
+                f"Extracted protein ID '{protein_id}' from '{filename}' (pattern 2)"
+            )
             return protein_id
 
         # Pattern 3: Just PROTEINID (no suffix)
@@ -104,7 +108,9 @@ class BoltzProcessor:
         match = re.match(pattern3, base_name)
         if match:
             protein_id = match.group(1)
-            logger.debug(f"Extracted protein ID '{protein_id}' from '{filename}' (pattern 3)")
+            logger.debug(
+                f"Extracted protein ID '{protein_id}' from '{filename}' (pattern 3)"
+            )
             return protein_id
 
         logger.debug(f"Could not extract protein ID from filename '{filename}'")
@@ -222,16 +228,24 @@ class BoltzProcessor:
             raise FileNotFoundError(f"Input file not found: {input_file}")
 
         # Enhance with MSA if available
-        enhanced_input_file, msa_found, original_yaml_data = self._enhance_yaml_with_msa(
-            input_file
+        enhanced_input_file, msa_found, original_yaml_data = (
+            self._enhance_yaml_with_msa(input_file)
         )
 
         # Determine output directory
+        # Boltz always creates boltz_results_{input_name} inside --out_dir
         input_base = os.path.splitext(os.path.basename(input_file))[0]
+        
         if output_dir is None:
+            # No output_dir specified, boltz creates in current directory
             expected_output_dir = f"boltz_results_{input_base}"
+            out_dir_arg = None
         else:
-            expected_output_dir = output_dir
+            # output_dir specified - use its parent for --out_dir
+            # and expect boltz_results_{input_base} inside it
+            parent_dir = os.path.dirname(output_dir)
+            expected_output_dir = os.path.join(parent_dir, f"boltz_results_{input_base}")
+            out_dir_arg = parent_dir if parent_dir else None
 
         logger.info(f"Running Boltz prediction for {input_file}")
         logger.info(f"Output directory: {expected_output_dir}")
@@ -239,6 +253,10 @@ class BoltzProcessor:
         # Build command
         cmd = ["boltz", "predict", input_file]
 
+        # Add output directory if specified
+        if out_dir_arg:
+            cmd.extend(["--out_dir", out_dir_arg])
+
         # Add cache directory
         cmd.extend(["--cache", self.cache_dir])
 
@@ -259,7 +277,9 @@ class BoltzProcessor:
         # Handle MSA server option
         if msa_found:
             if use_msa_server_in_opts:
-                additional_args = [arg for arg in additional_args if arg != "--use_msa_server"]
+                additional_args = [
+                    arg for arg in additional_args if arg != "--use_msa_server"
+                ]
                 logger.info("Removed --use_msa_server since local MSA was found")
         else:
             if not use_msa_server_in_opts:
@@ -270,6 +290,11 @@ class BoltzProcessor:
         if not num_workers_in_opts:
             cmd.extend(["--num_workers", str(self.num_workers)])
 
+        # Disable cuequivariance kernels - they require cuda-devel image
+        # which is much larger. The performance difference is modest.
+        # TODO: Consider switching to cuda-devel base image if perf is critical
+        cmd.append("--no_kernels")
+
         cmd.extend(additional_args)
 
         # Log and run command
@@ -305,7 +330,9 @@ class BoltzProcessor:
 
         # Copy input config to output directory
         try:
-            config_dest = os.path.join(expected_output_dir, os.path.basename(input_file))
+            config_dest = os.path.join(
+                expected_output_dir, os.path.basename(input_file)
+            )
             shutil.copy2(input_file, config_dest)
             logger.debug(f"Copied input config to results: {config_dest}")
         except Exception as e:
@@ -346,7 +373,9 @@ def main():
         input_files = sorted(input_dir.glob("*.yaml"))
 
         if index >= len(input_files):
-            logger.error(f"Index {index} out of range. Found {len(input_files)} input files.")
+            logger.error(
+                f"Index {index} out of range. Found {len(input_files)} input files."
+            )
             raise RuntimeError(f"Index {index} out of range")
 
         input_file = input_files[index]

dayhoff_tools/cli/batch/__init__.py

@@ -12,6 +12,7 @@ import click
 
 from .commands.boltz import boltz
 from .commands.cancel import cancel
+from .commands.clean import clean
 from .commands.embed_t5 import embed_t5
 from .commands.finalize import finalize
 from .commands.list_jobs import list_jobs
@@ -36,6 +37,7 @@ def batch_cli():
       finalize   Combine results and clean up
       local      Run a chunk locally for debugging
       list       List recent jobs
+      clean      Remove old completed job directories
 
     \b
     Embedding Pipelines:
@@ -77,6 +79,7 @@ batch_cli.add_command(retry)
 batch_cli.add_command(finalize)
 batch_cli.add_command(local)
 batch_cli.add_command(list_jobs, name="list")
+batch_cli.add_command(clean)
 
 # Register pipeline commands
 batch_cli.add_command(embed_t5, name="embed-t5")

dayhoff_tools/cli/batch/aws_batch.py

@@ -242,6 +242,64 @@ class BatchClient:
             failed=status_summary.get("FAILED", 0),
         )
 
+    def get_job_statuses_batch(self, job_ids: list[str]) -> dict[str, str]:
+        """Get status for multiple jobs in a single API call.
+
+        AWS Batch allows up to 100 job IDs per describe_jobs call.
+        This method handles batching for larger lists.
+
+        Args:
+            job_ids: List of AWS Batch job IDs
+
+        Returns:
+            Dictionary mapping job_id -> status string
+            Status will be one of: SUBMITTED, PENDING, RUNNABLE, STARTING,
+            RUNNING, SUCCEEDED, FAILED, or "UNKNOWN" if not found.
+            For array jobs, derives overall status from child statuses.
+        """
+        if not job_ids:
+            return {}
+
+        results = {}
+        batch_size = 100  # AWS Batch limit
+
+        for i in range(0, len(job_ids), batch_size):
+            batch = job_ids[i : i + batch_size]
+            try:
+                response = self.batch.describe_jobs(jobs=batch)
+                for job in response.get("jobs", []):
+                    job_id = job.get("jobId")
+                    status = job.get("status", "UNKNOWN")
+
+                    # For array jobs, derive overall status from children
+                    if "arrayProperties" in job:
+                        summary = job["arrayProperties"].get("statusSummary", {})
+                        total = job["arrayProperties"].get("size", 0)
+                        succeeded = summary.get("SUCCEEDED", 0)
+                        failed = summary.get("FAILED", 0)
+
+                        if succeeded + failed == total:
+                            # All children complete
+                            status = "SUCCEEDED" if failed == 0 else "FAILED"
+                        elif summary.get("RUNNING", 0) > 0:
+                            status = "RUNNING"
+                        elif summary.get("STARTING", 0) > 0:
+                            status = "STARTING"
+                        elif summary.get("RUNNABLE", 0) > 0:
+                            status = "RUNNABLE"
+                        elif summary.get("PENDING", 0) > 0:
+                            status = "PENDING"
+
+                    results[job_id] = status
+            except ClientError as e:
+                logger.warning(f"Failed to describe batch of jobs: {e}")
+                # Mark these as unknown
+                for job_id in batch:
+                    if job_id not in results:
+                        results[job_id] = "UNKNOWN"
+
+        return results
+
     def get_failed_indices(self, job_id: str) -> list[int]:
         """Get the array indices that failed for an array job.
 
@@ -256,9 +314,7 @@ class BatchClient:
         # List child jobs with FAILED status
         try:
             paginator = self.batch.get_paginator("list_jobs")
-            for page in paginator.paginate(
-                arrayJobId=job_id, jobStatus="FAILED"
-            ):
+            for page in paginator.paginate(arrayJobId=job_id, jobStatus="FAILED"):
                 for job_summary in page.get("jobSummaryList", []):
                     # Extract array index from job ID (format: jobId:index)
                     child_id = job_summary.get("jobId", "")
@@ -361,7 +417,9 @@ class BatchClient:
                 timestamp = event.get("timestamp", 0)
                 message = event.get("message", "")
                 # Format timestamp
-                dt = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(timestamp / 1000))
+                dt = time.strftime(
+                    "%Y-%m-%d %H:%M:%S", time.localtime(timestamp / 1000)
+                )
                 messages.append(f"[{dt}] {message}")
 
         except ClientError as e:

dayhoff_tools/cli/batch/commands/boltz.py

@@ -289,7 +289,9 @@ def _run_local_mode(input_path: Path):
         click.echo(click.style("Error: No YAML files found", fg="red"), err=True)
         raise SystemExit(1)
 
-    click.echo(f"Found {len(yaml_files)} YAML files, will process: {yaml_files[0].name}")
+    click.echo(
+        f"Found {len(yaml_files)} YAML files, will process: {yaml_files[0].name}"
+    )
 
     # Create a temporary job directory structure
     temp_job_dir = input_path / ".local_boltz_job"
@@ -311,14 +313,25 @@ def _run_local_mode(input_path: Path):
     click.echo()
 
     cmd = [
-        "docker", "run", "--rm",
-        "--gpus", "all",
-        "-v", "/primordial:/primordial",
-        "-v", f"{temp_job_dir}:{temp_job_dir}",
-        "-e", f"JOB_DIR={temp_job_dir}",
-        "-e", "AWS_BATCH_JOB_ARRAY_INDEX=0",
-        "-e", "BOLTZ_CACHE=/primordial/.cache/boltz",
-        "-e", "MSA_DIR=/primordial/.cache/msas",
+        "docker",
+        "run",
+        "--rm",
+        "--gpus",
+        "all",
+        "-v",
+        "/primordial:/primordial",
+        "-v",
+        f"{temp_job_dir}:{temp_job_dir}",
+        "-e",
+        f"JOB_DIR={temp_job_dir}",
+        "-e",
+        "AWS_BATCH_JOB_ARRAY_INDEX=0",
+        "-e",
+        "BOLTZ_CACHE=/primordial/.cache/boltz",
+        "-e",
+        "MSA_DIR=/primordial/.cache/msas",
+        "-e",
+        "BOLTZ_OPTIONS=--no_kernels",
         DEFAULT_IMAGE_URI,
     ]
 
@@ -329,13 +342,17 @@ def _run_local_mode(input_path: Path):
         result = subprocess.run(cmd)
         if result.returncode != 0:
             click.echo(
-                click.style(f"Container exited with code {result.returncode}", fg="red"),
+                click.style(
+                    f"Container exited with code {result.returncode}", fg="red"
+                ),
                 err=True,
             )
             raise SystemExit(result.returncode)
 
         # Check for output
-        output_dirs = list(temp_output_dir.iterdir()) if temp_output_dir.exists() else []
+        output_dirs = (
+            list(temp_output_dir.iterdir()) if temp_output_dir.exists() else []
+        )
         if output_dirs:
             click.echo()
             click.echo(click.style("✓ Prediction complete!", fg="green"))
@@ -347,7 +364,9 @@ def _run_local_mode(input_path: Path):
 
     except FileNotFoundError:
         click.echo(
-            click.style("Error: Docker not found. Is Docker installed and running?", fg="red"),
+            click.style(
+                "Error: Docker not found. Is Docker installed and running?", fg="red"
+            ),
             err=True,
         )
         raise SystemExit(1)

dayhoff_tools/cli/batch/commands/cancel.py

@@ -33,8 +33,14 @@ def cancel(job_id, force, base_path):
         raise SystemExit(1)
 
     # Check if job can be cancelled
-    if manifest.status in (JobStatus.SUCCEEDED, JobStatus.FINALIZED, JobStatus.CANCELLED):
-        click.echo(f"Job {job_id} is already {manifest.status.value}, cannot cancel.", err=True)
+    if manifest.status in (
+        JobStatus.SUCCEEDED,
+        JobStatus.FINALIZED,
+        JobStatus.CANCELLED,
+    ):
+        click.echo(
+            f"Job {job_id} is already {manifest.status.value}, cannot cancel.", err=True
+        )
         raise SystemExit(1)
 
     # Get Batch job ID
@@ -53,10 +59,14 @@ def cancel(job_id, force, base_path):
 
         if force:
             click.echo(f"Terminating job {batch_job_id}...")
-            client.terminate_job(batch_job_id, reason="Terminated by user via dh batch cancel --force")
+            client.terminate_job(
+                batch_job_id, reason="Terminated by user via dh batch cancel --force"
+            )
         else:
             click.echo(f"Cancelling job {batch_job_id}...")
-            client.cancel_job(batch_job_id, reason="Cancelled by user via dh batch cancel")
+            client.cancel_job(
+                batch_job_id, reason="Cancelled by user via dh batch cancel"
+            )
 
         # Update manifest
         manifest.status = JobStatus.CANCELLED
@@ -70,9 +80,13 @@ def cancel(job_id, force, base_path):
             if retry_info.batch_job_id:
                 try:
                     if force:
-                        client.terminate_job(retry_info.batch_job_id, reason="Parent job cancelled")
+                        client.terminate_job(
+                            retry_info.batch_job_id, reason="Parent job cancelled"
+                        )
                     else:
-                        client.cancel_job(retry_info.batch_job_id, reason="Parent job cancelled")
+                        client.cancel_job(
+                            retry_info.batch_job_id, reason="Parent job cancelled"
+                        )
                     click.echo(f"  Also cancelled retry job: {retry_info.retry_id}")
                 except BatchError:
                     pass  # Retry job may already be complete

dayhoff_tools/cli/batch/commands/clean.py

@@ -0,0 +1,139 @@
+"""Clean command for removing old job directories."""
+
+import click
+
+from ..aws_batch import BatchClient, BatchError
+from ..manifest import (
+    BATCH_JOBS_BASE,
+    JobStatus,
+    delete_job_directory,
+    list_jobs as list_manifests,
+)
+from .status import format_time_ago, _aws_status_to_job_status
+
+
+@click.command("clean")
+@click.option("--user", help="Only clean jobs for this user")
+@click.option(
+    "--older-than",
+    type=int,
+    default=7,
+    help="Only clean jobs older than N days [default: 7]",
+)
+@click.option("--dry-run", is_flag=True, help="Show what would be cleaned without deleting")
+@click.option("--force", is_flag=True, help="Delete without confirmation")
+@click.option("--base-path", default=BATCH_JOBS_BASE, help="Base path for job data")
+def clean(user, older_than, dry_run, force, base_path):
+    """Remove completed job directories to free up space.
+
+    Only removes jobs that have SUCCEEDED or FAILED in AWS Batch.
+    Jobs that are still running or pending are never removed.
+
+    \b
+    Examples:
+      dh batch clean                     # Clean jobs older than 7 days
+      dh batch clean --older-than 1      # Clean jobs older than 1 day
+      dh batch clean --dry-run           # Show what would be cleaned
+      dh batch clean --user dma          # Only clean dma's jobs
+    """
+    from datetime import datetime, timedelta, timezone
+
+    cutoff = datetime.now(timezone.utc) - timedelta(days=older_than)
+
+    # Get all manifests
+    manifests = list_manifests(
+        base_path=base_path,
+        user=user,
+        status=None,
+        pipeline=None,
+        limit=500,
+    )
+
+    if not manifests:
+        click.echo("No jobs found.")
+        return
+
+    # Filter to old jobs
+    old_manifests = []
+    for m in manifests:
+        created = m.created
+        if created.tzinfo is None:
+            created = created.replace(tzinfo=timezone.utc)
+        if created < cutoff:
+            old_manifests.append(m)
+
+    if not old_manifests:
+        click.echo(f"No jobs older than {older_than} days found.")
+        return
+
+    # Get live statuses for old jobs
+    batch_job_ids = []
+    manifest_to_batch_id = {}
+    for m in old_manifests:
+        if m.batch and m.batch.job_id:
+            batch_job_ids.append(m.batch.job_id)
+            manifest_to_batch_id[m.job_id] = m.batch.job_id
+
+    live_statuses = {}
+    if batch_job_ids:
+        try:
+            client = BatchClient()
+            live_statuses = client.get_job_statuses_batch(batch_job_ids)
+        except BatchError as e:
+            click.echo(f"Error: Could not fetch status from AWS Batch: {e}")
+            click.echo("Cannot safely clean jobs without knowing their status.")
+            return
+
+    # Find jobs that are safe to clean (SUCCEEDED or FAILED)
+    safe_to_clean = []
+    for manifest in old_manifests:
+        if manifest.job_id in manifest_to_batch_id:
+            batch_id = manifest_to_batch_id[manifest.job_id]
+            aws_status = live_statuses.get(batch_id, "UNKNOWN")
+            if aws_status in ("SUCCEEDED", "FAILED"):
+                safe_to_clean.append((manifest, aws_status))
+        elif manifest.status in (JobStatus.FINALIZED, JobStatus.CANCELLED):
+            # Already finalized or cancelled - safe to clean
+            safe_to_clean.append((manifest, manifest.status.value.upper()))
+
+    if not safe_to_clean:
+        click.echo(f"No completed jobs older than {older_than} days to clean.")
+        return
+
+    # Show what will be cleaned
+    click.echo()
+    click.echo(f"{'JOB ID':<35} {'STATUS':<12} {'CREATED':<12}")
+    click.echo("-" * 65)
+
+    for manifest, status in safe_to_clean:
+        click.echo(
+            f"{manifest.job_id:<35} "
+            f"{status:<12} "
+            f"{format_time_ago(manifest.created):<12}"
+        )
+
+    click.echo()
+    click.echo(f"Found {len(safe_to_clean)} completed jobs to clean.")
+
+    if dry_run:
+        click.echo("(dry-run: no changes made)")
+        return
+
+    # Confirm before deleting
+    if not force:
+        if not click.confirm("Delete these job directories?"):
+            click.echo("Cancelled.")
+            return
+
+    # Delete job directories
+    deleted = 0
+    for manifest, _ in safe_to_clean:
+        try:
+            delete_job_directory(manifest.job_id, base_path)
+            deleted += 1
+            click.echo(f"  Deleted: {manifest.job_id}")
+        except Exception as e:
+            click.echo(f"  Failed to delete {manifest.job_id}: {e}")
+
+    click.echo()
+    click.echo(f"Cleaned {deleted} job directories.")

dayhoff_tools/cli/batch/commands/embed_t5.py

@@ -32,14 +32,42 @@ DEFAULT_IMAGE_URI = "074735440724.dkr.ecr.us-east-1.amazonaws.com/dayhoff:embed-
 
 @click.command()
 @click.argument("input_fasta", type=click.Path(exists=True))
-@click.option("--workers", default=DEFAULT_WORKERS, type=int, help=f"Number of parallel workers [default: {DEFAULT_WORKERS}]")
-@click.option("--queue", default=DEFAULT_QUEUE, help=f"Batch queue [default: {DEFAULT_QUEUE}]")
-@click.option("--seqs-per-chunk", default=DEFAULT_SEQS_PER_CHUNK, type=int, help=f"Sequences per chunk [default: {DEFAULT_SEQS_PER_CHUNK}]")
-@click.option("--local", "run_local", is_flag=True, help="Run single chunk locally instead of Batch")
-@click.option("--shell", "run_shell", is_flag=True, help="Drop into container shell for debugging")
+@click.option(
+    "--workers",
+    default=DEFAULT_WORKERS,
+    type=int,
+    help=f"Number of parallel workers [default: {DEFAULT_WORKERS}]",
+)
+@click.option(
+    "--queue", default=DEFAULT_QUEUE, help=f"Batch queue [default: {DEFAULT_QUEUE}]"
+)
+@click.option(
+    "--seqs-per-chunk",
+    default=DEFAULT_SEQS_PER_CHUNK,
+    type=int,
+    help=f"Sequences per chunk [default: {DEFAULT_SEQS_PER_CHUNK}]",
+)
+@click.option(
+    "--local",
+    "run_local",
+    is_flag=True,
+    help="Run single chunk locally instead of Batch",
+)
+@click.option(
+    "--shell", "run_shell", is_flag=True, help="Drop into container shell for debugging"
+)
 @click.option("--dry-run", is_flag=True, help="Show plan without submitting")
 @click.option("--base-path", default=BATCH_JOBS_BASE, help="Base path for job data")
-def embed_t5(input_fasta, workers, queue, seqs_per_chunk, run_local, run_shell, dry_run, base_path):
+def embed_t5(
+    input_fasta,
+    workers,
+    queue,
+    seqs_per_chunk,
+    run_local,
+    run_shell,
+    dry_run,
+    base_path,
+):
     """Generate T5 protein embeddings for a FASTA file.
 
     Splits the input FASTA into chunks and processes them in parallel using
@@ -115,7 +143,14 @@ def _split_fasta(input_path: Path, output_dir: Path, seqs_per_chunk: int) -> int
     return num_chunks
 
 
-def _submit_batch_job(input_path: Path, workers: int, queue: str, seqs_per_chunk: int, dry_run: bool, base_path: str):
+def _submit_batch_job(
+    input_path: Path,
+    workers: int,
+    queue: str,
+    seqs_per_chunk: int,
+    dry_run: bool,
+    base_path: str,
+):
     """Submit embedding job to AWS Batch."""
     # Count sequences
     click.echo(f"Counting sequences in {input_path}...")
@@ -123,7 +158,9 @@ def _submit_batch_job(input_path: Path, workers: int, queue: str, seqs_per_chunk
     click.echo(f"Found {num_sequences:,} sequences")
 
     if num_sequences == 0:
-        click.echo(click.style("Error: No sequences found in input file", fg="red"), err=True)
+        click.echo(
+            click.style("Error: No sequences found in input file", fg="red"), err=True
+        )
         raise SystemExit(1)
 
     # Calculate chunks
@@ -223,7 +260,9 @@ def _submit_batch_job(input_path: Path, workers: int, queue: str, seqs_per_chunk
         click.echo(f"  Cancel:        dh batch cancel {job_id}")
         click.echo()
         click.echo("After completion:")
-        click.echo(f"  Finalize:      dh batch finalize {job_id} --output /primordial/embeddings.h5")
+        click.echo(
+            f"  Finalize:      dh batch finalize {job_id} --output /primordial/embeddings.h5"
+        )
 
     except BatchError as e:
         manifest.status = JobStatus.FAILED
@@ -265,12 +304,19 @@ def _run_local_mode(input_path: Path):
     click.echo()
 
     cmd = [
-        "docker", "run", "--rm",
-        "--gpus", "all",
-        "-v", "/primordial:/primordial",
-        "-v", f"{temp_job_dir}:{temp_job_dir}",
-        "-e", f"JOB_DIR={temp_job_dir}",
-        "-e", "AWS_BATCH_JOB_ARRAY_INDEX=0",
+        "docker",
+        "run",
+        "--rm",
+        "--gpus",
+        "all",
+        "-v",
+        "/primordial:/primordial",
+        "-v",
+        f"{temp_job_dir}:{temp_job_dir}",
+        "-e",
+        f"JOB_DIR={temp_job_dir}",
+        "-e",
+        "AWS_BATCH_JOB_ARRAY_INDEX=0",
         DEFAULT_IMAGE_URI,
     ]
 
@@ -281,7 +327,9 @@ def _run_local_mode(input_path: Path):
         result = subprocess.run(cmd)
         if result.returncode != 0:
             click.echo(
-                click.style(f"Container exited with code {result.returncode}", fg="red"),
+                click.style(
+                    f"Container exited with code {result.returncode}", fg="red"
+                ),
                 err=True,
             )
             raise SystemExit(result.returncode)
@@ -303,7 +351,9 @@ def _run_local_mode(input_path: Path):
 
     except FileNotFoundError:
         click.echo(
-            click.style("Error: Docker not found. Is Docker installed and running?", fg="red"),
+            click.style(
+                "Error: Docker not found. Is Docker installed and running?", fg="red"
+            ),
             err=True,
         )
         raise SystemExit(1)
@@ -318,13 +368,22 @@ def _run_shell_mode(input_path: Path):
     input_dir = input_path.parent
 
     cmd = [
-        "docker", "run", "--rm", "-it",
-        "--gpus", "all",
-        "-v", "/primordial:/primordial",
-        "-v", f"{input_dir}:/input",
-        "-e", "JOB_DIR=/input",
-        "-e", "AWS_BATCH_JOB_ARRAY_INDEX=0",
-        "--entrypoint", "/bin/bash",
+        "docker",
+        "run",
+        "--rm",
+        "-it",
+        "--gpus",
+        "all",
+        "-v",
+        "/primordial:/primordial",
+        "-v",
+        f"{input_dir}:/input",
+        "-e",
+        "JOB_DIR=/input",
+        "-e",
+        "AWS_BATCH_JOB_ARRAY_INDEX=0",
+        "--entrypoint",
+        "/bin/bash",
         DEFAULT_IMAGE_URI,
     ]
 
@@ -335,7 +394,9 @@ def _run_shell_mode(input_path: Path):
         subprocess.run(cmd)
     except FileNotFoundError:
         click.echo(
-            click.style("Error: Docker not found. Is Docker installed and running?", fg="red"),
+            click.style(
+                "Error: Docker not found. Is Docker installed and running?", fg="red"
+            ),
             err=True,
         )
         raise SystemExit(1)

dayhoff_tools/cli/batch/commands/finalize.py

@@ -17,9 +17,18 @@ from ..manifest import (
 
 @click.command()
 @click.argument("job_id")
-@click.option("--output", required=True, type=click.Path(), help="Output path for combined results")
+@click.option(
+    "--output",
+    required=True,
+    type=click.Path(),
+    help="Output path for combined results",
+)
 @click.option("--force", is_flag=True, help="Finalize even if some chunks failed")
-@click.option("--keep-intermediates", is_flag=True, help="Don't delete job directory after finalizing")
+@click.option(
+    "--keep-intermediates",
+    is_flag=True,
+    help="Don't delete job directory after finalizing",
+)
 @click.option("--base-path", default=BATCH_JOBS_BASE, help="Base path for job data")
 def finalize(job_id, output, force, keep_intermediates, base_path):
     """Combine results and clean up job intermediates.
@@ -59,7 +68,9 @@ def finalize(job_id, output, force, keep_intermediates, base_path):
             click.echo(f"  dh batch retry {job_id}")
             raise SystemExit(1)
         click.echo()
-        click.echo(click.style("Warning: Finalizing with incomplete chunks", fg="yellow"))
+        click.echo(
+            click.style("Warning: Finalizing with incomplete chunks", fg="yellow")
+        )
 
     # Update status
     manifest.status = JobStatus.FINALIZING
@@ -132,24 +143,38 @@ def _finalize_embeddings(output_dir: Path, output_path: Path):
     output_path.parent.mkdir(parents=True, exist_ok=True)
 
     try:
-        from dayhoff_tools.h5 import combine_h5_files, deduplicate_h5_file, optimize_protein_embedding_chunks
-
-        # Combine H5 files
-        click.echo("Combining H5 files...")
-        # Get list of h5 file paths as strings
-        h5_file_paths = [str(f) for f in h5_files]
-        combine_h5_files(
-            input_files=h5_file_paths,
-            output_file=str(output_path),
+        from dayhoff_tools.h5 import (
+            combine_h5_files,
+            deduplicate_h5_file,
+            optimize_protein_embedding_chunks,
         )
+        import tempfile
+
+        if len(h5_files) == 1:
+            # Single file - just copy, no need to combine/dedup/optimize
+            click.echo("Single chunk - copying directly...")
+            shutil.copy2(h5_files[0], output_path)
+        else:
+            # Multiple files - combine, deduplicate, and optimize
+            with tempfile.TemporaryDirectory() as tmpdir:
+                combined_path = Path(tmpdir) / "combined.h5"
+                deduped_path = Path(tmpdir) / "deduped.h5"
+
+                # Combine H5 files
+                click.echo("Combining H5 files...")
+                h5_file_paths = [str(f) for f in h5_files]
+                combine_h5_files(
+                    input_files=h5_file_paths,
+                    output_file=str(combined_path),
+                )
 
-        # Deduplicate
-        click.echo("Deduplicating...")
-        deduplicate_h5_file(str(output_path))
+                # Deduplicate
+                click.echo("Deduplicating...")
+                deduplicate_h5_file(str(combined_path), str(deduped_path))
 
-        # Optimize chunks
-        click.echo("Optimizing chunks...")
-        optimize_protein_embedding_chunks(str(output_path))
+                # Optimize chunks
+                click.echo("Optimizing chunks...")
+                optimize_protein_embedding_chunks(str(deduped_path), str(output_path))
 
         click.echo(click.style("✓ H5 files combined successfully", fg="green"))
 

dayhoff_tools/cli/batch/commands/list_jobs.py

@@ -2,8 +2,9 @@
 
 import click
 
+from ..aws_batch import BatchClient, BatchError
 from ..manifest import BATCH_JOBS_BASE, JobStatus, list_jobs as list_manifests
-from .status import format_status, format_time_ago
+from .status import format_status, format_time_ago, _aws_status_to_job_status
 
 
 @click.command("list")
@@ -15,12 +16,15 @@ from .status import format_status, format_time_ago
     help="Filter by status",
 )
 @click.option("--pipeline", help="Filter by pipeline type")
-@click.option("--limit", default=20, type=int, help="Maximum number of jobs to show [default: 20]")
+@click.option(
+    "--limit", default=20, type=int, help="Maximum number of jobs to show [default: 20]"
+)
 @click.option("--base-path", default=BATCH_JOBS_BASE, help="Base path for job data")
 def list_jobs(user, status_filter, pipeline, limit, base_path):
     """List recent batch jobs.
 
     Shows a table of recent jobs with their status, pipeline type, and creation time.
+    Status is fetched live from AWS Batch.
 
     \b
     Examples:
@@ -32,12 +36,13 @@ def list_jobs(user, status_filter, pipeline, limit, base_path):
     """
     status_enum = JobStatus(status_filter) if status_filter else None
 
+    # Fetch more manifests than requested to allow filtering by live status
     manifests = list_manifests(
         base_path=base_path,
         user=user,
-        status=status_enum,
+        status=None,  # Don't filter by status yet - will filter after getting live status
         pipeline=pipeline,
-        limit=limit,
+        limit=limit * 3,  # Fetch extra to account for status filtering
     )
 
     if not manifests:
@@ -46,6 +51,51 @@ def list_jobs(user, status_filter, pipeline, limit, base_path):
             click.echo("Try removing filters to see all jobs.")
         return
 
+    # Collect AWS Batch job IDs for live status lookup
+    batch_job_ids = []
+    manifest_to_batch_id = {}
+    for m in manifests:
+        if m.batch and m.batch.job_id:
+            batch_job_ids.append(m.batch.job_id)
+            manifest_to_batch_id[m.job_id] = m.batch.job_id
+
+    # Fetch live statuses from AWS Batch
+    live_statuses = {}
+    if batch_job_ids:
+        try:
+            client = BatchClient()
+            live_statuses = client.get_job_statuses_batch(batch_job_ids)
+        except BatchError as e:
+            click.echo(f"Warning: Could not fetch live status from AWS Batch: {e}")
+
+    # Build display data with live status
+    display_data = []
+    for manifest in manifests:
+        # Use live status if available, otherwise fall back to manifest status
+        if manifest.job_id in manifest_to_batch_id:
+            batch_id = manifest_to_batch_id[manifest.job_id]
+            aws_status = live_statuses.get(batch_id)
+            if aws_status:
+                live_status = _aws_status_to_job_status(aws_status)
+            else:
+                live_status = manifest.status
+        else:
+            live_status = manifest.status
+
+        # Apply status filter if specified
+        if status_enum and live_status != status_enum:
+            continue
+
+        display_data.append((manifest, live_status))
+
+        # Stop once we have enough
+        if len(display_data) >= limit:
+            break
+
+    if not display_data:
+        click.echo("No jobs found matching filters.")
+        return
+
     # Print header
     click.echo()
     click.echo(
@@ -53,17 +103,17 @@ def list_jobs(user, status_filter, pipeline, limit, base_path):
     )
     click.echo("-" * 85)
 
-    for manifest in manifests:
+    for manifest, live_status in display_data:
         click.echo(
             f"{manifest.job_id:<35} "
-            f"{format_status(manifest.status):<21} "  # Extra space for ANSI color codes
+            f"{format_status(live_status):<21} "  # Extra space for ANSI color codes
             f"{manifest.pipeline:<12} "
             f"{manifest.user:<10} "
             f"{format_time_ago(manifest.created):<12}"
         )
 
     click.echo()
-    click.echo(f"Showing {len(manifests)} jobs.")
+    click.echo(f"Showing {len(display_data)} jobs.")
 
     # Show filter hints
     hints = []

dayhoff_tools/cli/batch/commands/local.py

@@ -10,7 +10,12 @@ from ..manifest import BATCH_JOBS_BASE, get_job_dir, load_manifest
 @click.command()
 @click.argument("job_id")
 @click.option("--index", required=True, type=int, help="Array index to run")
-@click.option("--shell", "run_shell", is_flag=True, help="Drop into shell instead of running command")
+@click.option(
+    "--shell",
+    "run_shell",
+    is_flag=True,
+    help="Drop into shell instead of running command",
+)
 @click.option("--base-path", default=BATCH_JOBS_BASE, help="Base path for job data")
 def local(job_id, index, run_shell, base_path):
     """Run a job chunk locally for debugging.
@@ -54,13 +59,21 @@ def local(job_id, index, run_shell, base_path):
 
     # Build Docker command
     cmd = [
-        "docker", "run", "--rm",
-        "--gpus", "all",
-        "-v", "/primordial:/primordial",
-        "-v", f"{job_dir}:{job_dir}",
-        "-e", f"AWS_BATCH_JOB_ARRAY_INDEX={index}",
-        "-e", f"JOB_DIR={job_dir}",
-        "-e", f"JOB_ID={job_id}",
+        "docker",
+        "run",
+        "--rm",
+        "--gpus",
+        "all",
+        "-v",
+        "/primordial:/primordial",
+        "-v",
+        f"{job_dir}:{job_dir}",
+        "-e",
+        f"AWS_BATCH_JOB_ARRAY_INDEX={index}",
+        "-e",
+        f"JOB_DIR={job_dir}",
+        "-e",
+        f"JOB_ID={job_id}",
     ]
 
     if run_shell:
@@ -81,7 +94,9 @@ def local(job_id, index, run_shell, base_path):
         result = subprocess.run(cmd)
         if result.returncode != 0:
             click.echo(
-                click.style(f"Container exited with code {result.returncode}", fg="red"),
+                click.style(
+                    f"Container exited with code {result.returncode}", fg="red"
+                ),
                 err=True,
             )
             raise SystemExit(result.returncode)
@@ -89,7 +104,9 @@ def local(job_id, index, run_shell, base_path):
             click.echo(click.style("✓ Container completed successfully", fg="green"))
     except FileNotFoundError:
         click.echo(
-            click.style("Error: Docker not found. Is Docker installed and running?", fg="red"),
+            click.style(
+                "Error: Docker not found. Is Docker installed and running?", fg="red"
+            ),
             err=True,
         )
         raise SystemExit(1)

dayhoff_tools/cli/batch/commands/logs.py

@@ -77,7 +77,9 @@ def _show_job_logs(client: BatchClient, batch_job_id: str, tail: int, follow: bo
                     click.echo(f"  dh batch logs {batch_job_id.split('-')[0]} --failed")
                     click.echo()
                     click.echo("To view logs for a specific index:")
-                    click.echo(f"  dh batch logs {batch_job_id.split('-')[0]} --index {failed_indices[0]}")
+                    click.echo(
+                        f"  dh batch logs {batch_job_id.split('-')[0]} --index {failed_indices[0]}"
+                    )
             return
 
         # Single job - show logs
@@ -94,7 +96,9 @@ def _show_job_logs(client: BatchClient, batch_job_id: str, tail: int, follow: bo
         click.echo(click.style(f"Error fetching logs: {e}", fg="red"), err=True)
 
 
-def _show_index_logs(client: BatchClient, batch_job_id: str, index: int, tail: int, follow: bool):
+def _show_index_logs(
+    client: BatchClient, batch_job_id: str, index: int, tail: int, follow: bool
+):
     """Show logs for a specific array index."""
     child_job_id = f"{batch_job_id}:{index}"
 

dayhoff_tools/cli/batch/commands/retry.py

@@ -19,7 +19,9 @@ from ..manifest import (
 @click.command()
 @click.argument("job_id")
 @click.option("--indices", help="Specific indices to retry (comma-separated)")
-@click.option("--dry-run", is_flag=True, help="Show what would be retried without submitting")
+@click.option(
+    "--dry-run", is_flag=True, help="Show what would be retried without submitting"
+)
 @click.option("--base-path", default=BATCH_JOBS_BASE, help="Base path for job data")
 def retry(job_id, indices, dry_run, base_path):
     """Retry failed chunks of a batch job.
@@ -112,7 +114,9 @@ def retry(job_id, indices, dry_run, base_path):
         click.echo(f"  View logs:     dh batch logs {job_id}")
 
     except BatchError as e:
-        click.echo(click.style(f"✗ Failed to submit retry job: {e}", fg="red"), err=True)
+        click.echo(
+            click.style(f"✗ Failed to submit retry job: {e}", fg="red"), err=True
+        )
         raise SystemExit(1)
 
 

dayhoff_tools/cli/batch/commands/status.py

@@ -3,7 +3,12 @@
 import click
 
 from ..aws_batch import BatchClient, BatchError
-from ..manifest import BATCH_JOBS_BASE, JobStatus, list_jobs as list_manifests, load_manifest
+from ..manifest import (
+    BATCH_JOBS_BASE,
+    JobStatus,
+    list_jobs as list_manifests,
+    load_manifest,
+)
 
 
 def format_status(status: JobStatus) -> str:
@@ -75,21 +80,79 @@ def status(job_id, user, status_filter, pipeline, base_path):
         _show_job_list(user, status_filter, pipeline, base_path)
 
 
+def _aws_status_to_job_status(aws_status: str) -> JobStatus:
+    """Convert AWS Batch status to JobStatus enum."""
+    mapping = {
+        "SUBMITTED": JobStatus.SUBMITTED,
+        "PENDING": JobStatus.PENDING,
+        "RUNNABLE": JobStatus.RUNNING,  # Runnable means waiting for compute
+        "STARTING": JobStatus.RUNNING,
+        "RUNNING": JobStatus.RUNNING,
+        "SUCCEEDED": JobStatus.SUCCEEDED,
+        "FAILED": JobStatus.FAILED,
+    }
+    return mapping.get(aws_status, JobStatus.SUBMITTED)
+
+
 def _show_job_list(user, status_filter, pipeline, base_path):
     """Show a list of recent jobs."""
     status_enum = JobStatus(status_filter) if status_filter else None
     manifests = list_manifests(
         base_path=base_path,
         user=user,
-        status=status_enum,
+        status=None,  # Don't filter yet - we'll filter after getting live status
         pipeline=pipeline,
-        limit=20,
+        limit=50,  # Fetch more, filter later
     )
 
     if not manifests:
         click.echo("No jobs found.")
         return
 
+    # Collect AWS Batch job IDs for live status lookup
+    batch_job_ids = []
+    manifest_to_batch_id = {}
+    for m in manifests:
+        if m.batch and m.batch.job_id:
+            batch_job_ids.append(m.batch.job_id)
+            manifest_to_batch_id[m.job_id] = m.batch.job_id
+
+    # Fetch live statuses from AWS Batch
+    live_statuses = {}
+    if batch_job_ids:
+        try:
+            client = BatchClient()
+            live_statuses = client.get_job_statuses_batch(batch_job_ids)
+        except BatchError as e:
+            click.echo(f"Warning: Could not fetch live status from AWS Batch: {e}")
+
+    # Build display data with live status
+    display_data = []
+    for manifest in manifests:
+        # Use live status if available, otherwise fall back to manifest status
+        if manifest.job_id in manifest_to_batch_id:
+            batch_id = manifest_to_batch_id[manifest.job_id]
+            aws_status = live_statuses.get(batch_id)
+            if aws_status:
+                live_status = _aws_status_to_job_status(aws_status)
+            else:
+                live_status = manifest.status
+        else:
+            live_status = manifest.status
+
+        # Apply status filter if specified
+        if status_enum and live_status != status_enum:
+            continue
+
+        display_data.append((manifest, live_status))
+
+    if not display_data:
+        click.echo("No jobs found matching filters.")
+        return
+
+    # Limit to 20 after filtering
+    display_data = display_data[:20]
+
     # Print header
     click.echo()
     click.echo(
@@ -97,17 +160,17 @@ def _show_job_list(user, status_filter, pipeline, base_path):
     )
     click.echo("-" * 85)
 
-    for manifest in manifests:
+    for manifest, live_status in display_data:
         click.echo(
             f"{manifest.job_id:<35} "
-            f"{format_status(manifest.status):<21} "  # Extra space for color codes
+            f"{format_status(live_status):<21} "  # Extra space for color codes
             f"{manifest.pipeline:<12} "
             f"{manifest.user:<10} "
             f"{format_time_ago(manifest.created):<12}"
         )
 
     click.echo()
-    click.echo(f"Showing {len(manifests)} most recent jobs.")
+    click.echo(f"Showing {len(display_data)} most recent jobs.")
     click.echo("Use 'dh batch status <job-id>' for details.")
 
 
@@ -125,8 +188,12 @@ def _show_job_details(job_id: str, base_path: str):
     click.echo(f"Status:    {format_status(manifest.status)}")
     click.echo(f"Pipeline:  {manifest.pipeline}")
     click.echo(f"User:      {manifest.user}")
-    click.echo(f"Created:   {manifest.created.isoformat()} ({format_time_ago(manifest.created)})")
-    click.echo(f"Updated:   {manifest.updated.isoformat()} ({format_time_ago(manifest.updated)})")
+    click.echo(
+        f"Created:   {manifest.created.isoformat()} ({format_time_ago(manifest.created)})"
+    )
+    click.echo(
+        f"Updated:   {manifest.updated.isoformat()} ({format_time_ago(manifest.updated)})"
+    )
 
     if manifest.input:
         click.echo()
@@ -182,7 +249,9 @@ def _show_job_details(job_id: str, base_path: str):
         click.echo(f"  Retry:       dh batch retry {job_id}")
     elif manifest.status == JobStatus.SUCCEEDED:
         click.echo("Next steps:")
-        click.echo(f"  Finalize:    dh batch finalize {job_id} --output /primordial/output.h5")
+        click.echo(
+            f"  Finalize:    dh batch finalize {job_id} --output /primordial/output.h5"
+        )
 
 
 def _show_array_status(batch_job_id: str):
@@ -205,10 +274,14 @@ def _show_array_status(batch_job_id: str):
         if array_status.is_complete:
             pct = array_status.success_rate * 100
             color = "green" if pct == 100 else "yellow" if pct > 90 else "red"
-            click.echo(f"    Complete:  {click.style(f'{pct:.1f}%', fg=color)} success rate")
+            click.echo(
+                f"    Complete:  {click.style(f'{pct:.1f}%', fg=color)} success rate"
+            )
         else:
             pct = array_status.completed / array_status.total * 100
-            click.echo(f"    Progress:  {pct:.1f}% ({array_status.completed}/{array_status.total})")
+            click.echo(
+                f"    Progress:  {pct:.1f}% ({array_status.completed}/{array_status.total})"
+            )
 
     except BatchError as e:
         click.echo(f"    (Could not fetch live status: {e})")

dayhoff_tools/cli/batch/commands/submit.py

@@ -25,9 +25,13 @@ DEFAULT_QUEUE = "t4-1x-spot"
 
 
 @click.command()
-@click.option("-f", "--file", "config_file", type=click.Path(exists=True), help="Config file path")
+@click.option(
+    "-f", "--file", "config_file", type=click.Path(exists=True), help="Config file path"
+)
 @click.option("--command", help="Command to run (alternative to config file)")
-@click.option("--queue", default=DEFAULT_QUEUE, help=f"Batch queue [default: {DEFAULT_QUEUE}]")
+@click.option(
+    "--queue", default=DEFAULT_QUEUE, help=f"Batch queue [default: {DEFAULT_QUEUE}]"
+)
 @click.option("--memory", default="30G", help="Memory limit (e.g., 30G)")
 @click.option("--vcpus", default=8, type=int, help="Number of vCPUs")
 @click.option("--gpus", default=1, type=int, help="Number of GPUs")
@@ -91,7 +95,9 @@ def submit(
     # Override with command-line options
     job_command = command or config.get("command")
     if not job_command:
-        raise click.UsageError("Must specify --command or provide config file with 'command' field")
+        raise click.UsageError(
+            "Must specify --command or provide config file with 'command' field"
+        )
 
     job_queue = queue if queue != DEFAULT_QUEUE else config.get("queue", queue)
     job_memory = memory if memory != "30G" else config.get("memory", memory)

dayhoff_tools/cli/batch/manifest.py

@@ -33,7 +33,9 @@ class InputConfig(BaseModel):
     """Configuration for job input."""
 
     source: str = Field(..., description="Path to input file or directory")
-    num_sequences: int | None = Field(None, description="Number of sequences (for FASTA)")
+    num_sequences: int | None = Field(
+        None, description="Number of sequences (for FASTA)"
+    )
     num_chunks: int | None = Field(None, description="Number of chunks created")
     sequences_per_chunk: int | None = Field(None, description="Sequences per chunk")
 

dayhoff_tools/cli/main.py

@@ -6,7 +6,10 @@ from importlib.metadata import PackageNotFoundError, version
 import typer
 from dayhoff_tools.cli.cloud_commands import aws_app, gcp_app
 from dayhoff_tools.cli.github_commands import gh_app
-from dayhoff_tools.cli.engine1 import engine_app as engine1_app, studio_app as studio1_app
+from dayhoff_tools.cli.engine1 import (
+    engine_app as engine1_app,
+    studio_app as studio1_app,
+)
 from dayhoff_tools.cli.utility_commands import (
     add_dependency,
     build_and_upload_wheel,
@@ -70,6 +73,7 @@ app.add_typer(gcp_app, name="gcp", help="Manage GCP authentication and impersona
 app.add_typer(aws_app, name="aws", help="Manage AWS SSO authentication.")
 app.add_typer(gh_app, name="gh", help="Manage GitHub authentication.")
 
+
 # Engine and Studio commands (v2 - new default with progress tracking)
 # These use Click instead of Typer, so we need a passthrough wrapper
 @app.command(

{dayhoff_tools-1.14.6.dist-info → dayhoff_tools-1.14.8.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dayhoff-tools
-Version: 1.14.6
+Version: 1.14.8
 Summary: Common tools for all the repos at Dayhoff Labs
 Author: Daniel Martin-Alarcon
 Author-email: dma@dayhofflabs.com

{dayhoff_tools-1.14.6.dist-info → dayhoff_tools-1.14.8.dist-info}/RECORD

@@ -2,26 +2,27 @@ dayhoff_tools/__init__.py,sha256=M5zThPyEBRYa5CfwlzKhcqTevWn3OKu62cjV6Zqie2A,469
 dayhoff_tools/batch/__init__.py,sha256=N7v1pUg3rp68W6J7cX2Gssxdfp57Z7G0WeFJh7gKeiM,163
 dayhoff_tools/batch/workers/__init__.py,sha256=DJTtiBSE5k9J6qwG_MsXDn8hgz_lvEdaMwqKa6YyqXo,369
 dayhoff_tools/batch/workers/base.py,sha256=Jie5ScJrPSRkmLQdcM2hPJTmk3pIeLL0oA8Y9pkP0HA,4269
-dayhoff_tools/batch/workers/boltz.py,sha256=QTIqMWuV_o3v1tYaZhvjyCl5F65ZlPcc0-TXtitMaFQ,14663
+dayhoff_tools/batch/workers/boltz.py,sha256=mZ1YUgzNXUf_Mw97SvtqOqVV5ON7zSwrFsrBuLVWJlU,15708
 dayhoff_tools/batch/workers/embed_t5.py,sha256=A5WqsQa1WZ7_la5X5wt0XUP-VwOglH04WyEINhwipeY,2750
 dayhoff_tools/chemistry/standardizer.py,sha256=uMn7VwHnx02nc404eO6fRuS4rsl4dvSPf2ElfZDXEpY,11188
 dayhoff_tools/chemistry/utils.py,sha256=jt-7JgF-GeeVC421acX-bobKbLU_X94KNOW24p_P-_M,2257
 dayhoff_tools/cli/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-dayhoff_tools/cli/batch/__init__.py,sha256=gEuiDPWio8ihrIfF_Q6-5hKOnliPTxDKQrKN8-4Y3Ac,2320
-dayhoff_tools/cli/batch/aws_batch.py,sha256=L6lNpUaxyhMvLcM3dF8yEi2yCgTNdviaNMmG75bpBIc,12719
+dayhoff_tools/cli/batch/__init__.py,sha256=0QMxkelGUdvdd7-P-WG8YX2h0qjkKKb-0JbEHGrH7pE,2437
+dayhoff_tools/cli/batch/aws_batch.py,sha256=DOp8KvmTrie15O61DP1HT13PSr3s5imxM-VZHvaLO7Q,15174
 dayhoff_tools/cli/batch/commands/__init__.py,sha256=1xRzzL_mc1hz1Pv0OWNr-g6fkL5XbEsOTGHzrqddLCA,458
-dayhoff_tools/cli/batch/commands/boltz.py,sha256=WARoklK5QTo71echUkm-anyj2AlEI9AEtWaiD_ds-EU,11759
-dayhoff_tools/cli/batch/commands/cancel.py,sha256=5FHLRoq8nvOr8HuKxXUZqAyAdjQP8seaNSaeAdhOoE8,2890
-dayhoff_tools/cli/batch/commands/embed_t5.py,sha256=UtF8ulJcz15sdd1QzvPVPeioN_AMks1DZ7c6vKtwhDE,11052
-dayhoff_tools/cli/batch/commands/finalize.py,sha256=i9In09vrzc9jQlCA6F82ceuZB9mHUXeOaYYyRUrHQCc,7254
-dayhoff_tools/cli/batch/commands/list_jobs.py,sha256=y90_XqFI2qcvV2XUC01P3P2aIdtl5hXP3cBDEofsRtU,2432
-dayhoff_tools/cli/batch/commands/local.py,sha256=vOTojTAp4YZXIMLKNqc76oIE0oWMjB1mtTe6r19LrEc,3075
-dayhoff_tools/cli/batch/commands/logs.py,sha256=FOYpANQCsQTJKigq77k1xbhgOqtRPw9hyHkljILWlmo,5241
-dayhoff_tools/cli/batch/commands/retry.py,sha256=FQMi77oh61i-JOD77OmoJECw03aWQIVS-MizKggSu1g,4481
-dayhoff_tools/cli/batch/commands/status.py,sha256=BZyCF115XNqx4KXyLASsowY_e4R29BYQJyaK_EulPL8,7553
-dayhoff_tools/cli/batch/commands/submit.py,sha256=9Ur2PFkb5vbWv8Xebqs8dsGDh28uAkOQBaMuZsLgnec,6752
+dayhoff_tools/cli/batch/commands/boltz.py,sha256=KbBxeF0HS_ysgd7MyeJgrkcGTAXChJLQxOhvJ14SxjY,11989
+dayhoff_tools/cli/batch/commands/cancel.py,sha256=ZnHAJVzMGC0_1EQGpMSdYUlzm9yi-E9NxRJKBsetYW8,3111
+dayhoff_tools/cli/batch/commands/clean.py,sha256=v8R_YtsdkfPjrFmo7a9cMdbDkWxotSmUeE26dnRlsnU,4537
+dayhoff_tools/cli/batch/commands/embed_t5.py,sha256=QXFydAw0wndevdzXF1cxikxMmvn1BuQ5p9lwutQFajU,11453
+dayhoff_tools/cli/batch/commands/finalize.py,sha256=REt-1f1Vb7OycJJ6JTUV_xJQVKBPXz0RxxhEuNfd39g,7957
+dayhoff_tools/cli/batch/commands/list_jobs.py,sha256=COfxZddDVUAHeTayNAB3ruYNhgrE3osgFxY2qzf33cg,4284
+dayhoff_tools/cli/batch/commands/local.py,sha256=dZeKhNakaM1jS-EoByAwg1nWspRRoOmYzcwzjEKBaIA,3226
+dayhoff_tools/cli/batch/commands/logs.py,sha256=ctgJksdzFmqBdD18ePPsZe2BpuJYtHz2xAaMPnUplmQ,5293
+dayhoff_tools/cli/batch/commands/retry.py,sha256=PbstKr7iP7Uu4csisp3cvM8al9YOFw27XVlc591jLbE,4509
+dayhoff_tools/cli/batch/commands/status.py,sha256=k8ltke62vKUX_4MmnFDEd6H2s8KmVFA71UUXLbleA7o,9790
+dayhoff_tools/cli/batch/commands/submit.py,sha256=AXbvSReN8fLlzR5swE81pH7yvbCC1GUMCbsDrfoHAws,6786
 dayhoff_tools/cli/batch/job_id.py,sha256=mqr8SwcPlUWIYLaR_C4kACmL2ZFK8jaddd7B-45-XaQ,4246
-dayhoff_tools/cli/batch/manifest.py,sha256=2js-vbLkRipStN6_89T-wVUjgITA32EhN4eMdcw41hw,8610
+dayhoff_tools/cli/batch/manifest.py,sha256=blUw2ASxFTmdLV5ikz4VD91n5CN0V37rlyX11M0H2VE,8624
 dayhoff_tools/cli/cloud_commands.py,sha256=xrWQZm48e09GTh8RmYh3JqXUzW8mMy07dHk_77LqZx8,41199
 dayhoff_tools/cli/engine1/__init__.py,sha256=RE45X2IPCNUvQp6_OHI2lbjwoqexkH3roUPccIHkeJo,9540
 dayhoff_tools/cli/engine1/engine_core.py,sha256=TECO6766GCZTtbxOY7QRggwmjpMH9JZSuxjbwF46unU,27061
@@ -47,7 +48,7 @@ dayhoff_tools/cli/engines_studios/simulators/studio_status_simulator.py,sha256=6
 dayhoff_tools/cli/engines_studios/ssh_config.py,sha256=UCv-jf_zSuk7FUStkCQBAJz1QkxiSEwZbdSrwt_9SMU,2932
 dayhoff_tools/cli/engines_studios/studio_commands.py,sha256=KGSNZQS8MmM_DfQzT9SRZvuR3OK6NdIdOrqI2wJFyes,25984
 dayhoff_tools/cli/github_commands.py,sha256=pfrxI68LObGm_gtPlQN-gHPahHV4l9k9T4GqO99NNL0,8948
-dayhoff_tools/cli/main.py,sha256=6W-fdRfjkPGx4fnfAhmEKyFl6Ou1VOxy4IYFpcFWkrA,8489
+dayhoff_tools/cli/main.py,sha256=6ffnaFzui-bVd1ME7yThk_ZrMOofwStamEUkkYlminY,8503
 dayhoff_tools/cli/swarm_commands.py,sha256=5EyKj8yietvT5lfoz8Zx0iQvVaNgc3SJX1z2zQR6o6M,5614
 dayhoff_tools/cli/utility_commands.py,sha256=e2P4dCCtoqMUGNyb0lFBZ6GZpl5Zslm1qqE5qIvsy38,50765
 dayhoff_tools/deployment/base.py,sha256=uZnFvnPQx6pH_HmJbdThweAs3BrxMaDohpE3iX_-yk4,18377
@@ -70,7 +71,7 @@ dayhoff_tools/intake/uniprot.py,sha256=BZYJQF63OtPcBBnQ7_P9gulxzJtqyorgyuDiPeOJq
 dayhoff_tools/logs.py,sha256=DKdeP0k0kliRcilwvX0mUB2eipO5BdWUeHwh-VnsICs,838
 dayhoff_tools/sqlite.py,sha256=jV55ikF8VpTfeQqqlHSbY8OgfyfHj8zgHNpZjBLos_E,18672
 dayhoff_tools/warehouse.py,sha256=UETBtZD3r7WgvURqfGbyHlT7cxoiVq8isjzMuerKw8I,24475
-dayhoff_tools-1.14.6.dist-info/METADATA,sha256=3rnOAUIQQN2JYIy6MtfSgAraglhmDY64Aq7fQWs-iRo,3184
-dayhoff_tools-1.14.6.dist-info/WHEEL,sha256=zp0Cn7JsFoX2ATtOhtaFYIiE2rmFAD4OcMhtUki8W3U,88
-dayhoff_tools-1.14.6.dist-info/entry_points.txt,sha256=iAf4jteNqW3cJm6CO6czLxjW3vxYKsyGLZ8WGmxamSc,49
-dayhoff_tools-1.14.6.dist-info/RECORD,,
+dayhoff_tools-1.14.8.dist-info/METADATA,sha256=86xFQ0fq394JPcdZ-o2lRsWuDoLiGDkSw5SsePljptM,3184
+dayhoff_tools-1.14.8.dist-info/WHEEL,sha256=zp0Cn7JsFoX2ATtOhtaFYIiE2rmFAD4OcMhtUki8W3U,88
+dayhoff_tools-1.14.8.dist-info/entry_points.txt,sha256=iAf4jteNqW3cJm6CO6czLxjW3vxYKsyGLZ8WGmxamSc,49
+dayhoff_tools-1.14.8.dist-info/RECORD,,