up-cli 0.2.0__py3-none-any.whl → 0.5.0__py3-none-any.whl

up/commands/provenance.py

@@ -0,0 +1,267 @@
+"""up provenance - View AI operation history and lineage."""
+
+import json
+from pathlib import Path
+
+import click
+from rich.console import Console
+from rich.panel import Panel
+from rich.table import Table
+
+from up.core.provenance import get_provenance_manager, ProvenanceEntry
+
+console = Console()
+
+
+@click.group()
+def provenance():
+    """View AI operation history and lineage.
+    
+    Track the provenance of AI-generated code changes,
+    including models used, prompts, and verification results.
+    """
+    pass
+
+
+@provenance.command("show")
+@click.argument("entry_id", required=False)
+@click.option("--task", "-t", help="Show by task ID")
+def show_cmd(entry_id: str, task: str):
+    """Show details of a provenance entry.
+    
+    \b
+    Examples:
+      up provenance show abc123def456
+      up provenance show --task US-007
+    """
+    cwd = Path.cwd()
+    manager = get_provenance_manager(cwd)
+    
+    entry = None
+    if entry_id:
+        entry = manager.get_entry(entry_id)
+    elif task:
+        entry = manager.get_entry_for_task(task)
+    else:
+        console.print("[yellow]Provide an entry ID or --task[/]")
+        return
+    
+    if not entry:
+        console.print("[red]Entry not found[/]")
+        return
+    
+    _display_entry(entry)
+
+
+@provenance.command("list")
+@click.option("--limit", "-n", default=20, help="Number of entries to show")
+@click.option("--status", type=click.Choice(["pending", "accepted", "rejected"]), help="Filter by status")
+@click.option("--json", "as_json", is_flag=True, help="Output as JSON")
+def list_cmd(limit: int, status: str, as_json: bool):
+    """List recent provenance entries.
+    
+    \b
+    Examples:
+      up provenance list
+      up provenance list -n 50
+      up provenance list --status accepted
+    """
+    cwd = Path.cwd()
+    manager = get_provenance_manager(cwd)
+    
+    entries = manager.list_entries(limit=limit, status=status)
+    
+    if not entries:
+        console.print("[dim]No provenance entries found[/]")
+        console.print("Entries are created when running [cyan]up start[/]")
+        return
+    
+    if as_json:
+        output = [e.to_dict() for e in entries]
+        console.print(json.dumps(output, indent=2))
+        return
+    
+    table = Table(title=f"Provenance History (last {len(entries)})")
+    table.add_column("ID", style="cyan")
+    table.add_column("Task")
+    table.add_column("Model")
+    table.add_column("Status")
+    table.add_column("Files")
+    table.add_column("Tests")
+    table.add_column("Time")
+    
+    for entry in entries:
+        status_icon = {
+            "accepted": "🟢",
+            "rejected": "🔴",
+            "pending": "🟡",
+        }.get(entry.status, "⚪")
+        
+        tests_icon = "✓" if entry.tests_passed else "✗" if entry.tests_passed is False else "-"
+        
+        table.add_row(
+            entry.id,
+            entry.task_id[:10] if entry.task_id else "-",
+            entry.ai_model[:10],
+            f"{status_icon} {entry.status}",
+            str(len(entry.files_modified)),
+            tests_icon,
+            entry.created_at[:16]
+        )
+    
+    console.print(table)
+
+
+@provenance.command("stats")
+@click.option("--json", "as_json", is_flag=True, help="Output as JSON")
+def stats_cmd(as_json: bool):
+    """Show provenance statistics.
+    
+    Displays aggregated stats about AI operations,
+    including acceptance rates and test results.
+    """
+    cwd = Path.cwd()
+    manager = get_provenance_manager(cwd)
+    
+    stats = manager.get_stats()
+    
+    if as_json:
+        console.print(json.dumps(stats, indent=2))
+        return
+    
+    console.print(Panel.fit(
+        "[bold]AI Operation Statistics[/]",
+        border_style="blue"
+    ))
+    
+    # Overview
+    console.print("\n[bold]Overview[/]")
+    console.print(f"  Total Operations: {stats['total_operations']}")
+    console.print(f"  Accepted: [green]{stats['accepted']}[/]")
+    console.print(f"  Rejected: [red]{stats['rejected']}[/]")
+    console.print(f"  Pending: [yellow]{stats['pending']}[/]")
+    
+    if stats['total_operations'] > 0:
+        console.print(f"  Acceptance Rate: {stats['acceptance_rate']*100:.1f}%")
+    
+    # Code changes
+    console.print("\n[bold]Code Changes[/]")
+    console.print(f"  Lines Added: [green]+{stats['total_lines_added']}[/]")
+    console.print(f"  Lines Removed: [red]-{stats['total_lines_removed']}[/]")
+    
+    # Testing
+    if stats['tests_run'] > 0:
+        console.print("\n[bold]Testing[/]")
+        console.print(f"  Tests Run: {stats['tests_run']}")
+        console.print(f"  Tests Passed: {stats['tests_passed']}")
+        console.print(f"  Pass Rate: {stats['test_pass_rate']*100:.1f}%")
+    
+    # Models
+    if stats['models_used']:
+        console.print("\n[bold]AI Models Used[/]")
+        for model, count in sorted(stats['models_used'].items(), key=lambda x: -x[1]):
+            console.print(f"  {model}: {count}")
+
+
+@provenance.command("verify")
+@click.argument("entry_id")
+@click.option("--accept", "-a", is_flag=True, help="Mark as accepted")
+@click.option("--reject", "-r", is_flag=True, help="Mark as rejected")
+@click.option("--reason", help="Rejection reason")
+def verify_cmd(entry_id: str, accept: bool, reject: bool, reason: str):
+    """Verify a provenance entry.
+    
+    Mark an entry as accepted or rejected after review.
+    """
+    cwd = Path.cwd()
+    manager = get_provenance_manager(cwd)
+    
+    if not accept and not reject:
+        console.print("[yellow]Specify --accept or --reject[/]")
+        return
+    
+    entry = manager.get_entry(entry_id)
+    if not entry:
+        console.print(f"[red]Entry not found: {entry_id}[/]")
+        return
+    
+    if accept:
+        entry = manager.complete_operation(entry_id, status="accepted")
+        console.print(f"[green]✓[/] Entry {entry_id} marked as accepted")
+    elif reject:
+        entry = manager.reject_operation(entry_id, reason=reason or "Manual rejection")
+        console.print(f"[red]✗[/] Entry {entry_id} marked as rejected")
+        if reason:
+            console.print(f"  Reason: {reason}")
+
+
+def _display_entry(entry: ProvenanceEntry) -> None:
+    """Display a single provenance entry in detail."""
+    status_color = {
+        "accepted": "green",
+        "rejected": "red",
+        "pending": "yellow",
+    }.get(entry.status, "white")
+    
+    console.print(Panel.fit(
+        f"[bold]Provenance Entry[/] [{status_color}]{entry.status.upper()}[/]",
+        border_style=status_color
+    ))
+    
+    # Basic info
+    table = Table(show_header=False, box=None)
+    table.add_column("Key", style="dim")
+    table.add_column("Value")
+    
+    table.add_row("ID", f"[cyan]{entry.id}[/]")
+    table.add_row("Task", f"{entry.task_id} - {entry.task_title}")
+    table.add_row("AI Model", entry.ai_model)
+    table.add_row("Branch", entry.branch)
+    table.add_row("Commit", entry.commit_sha or "-")
+    table.add_row("Created", entry.created_at)
+    if entry.completed_at:
+        table.add_row("Completed", entry.completed_at)
+    
+    console.print(table)
+    
+    # Prompt preview
+    if entry.prompt_preview:
+        console.print("\n[bold]Prompt Preview[/]")
+        console.print(f"[dim]{entry.prompt_preview}[/]")
+    
+    # Files
+    if entry.files_modified:
+        console.print("\n[bold]Files Modified[/]")
+        for f in entry.files_modified[:10]:
+            console.print(f"  • {f}")
+        if len(entry.files_modified) > 10:
+            console.print(f"  ... and {len(entry.files_modified) - 10} more")
+    
+    # Context files
+    if entry.context_files:
+        console.print("\n[bold]Context Files[/]")
+        for f in entry.context_files[:5]:
+            console.print(f"  • {f}")
+    
+    # Verification
+    console.print("\n[bold]Verification[/]")
+    
+    def status_icon(val):
+        if val is True:
+            return "[green]✓[/]"
+        elif val is False:
+            return "[red]✗[/]"
+        return "[dim]-[/]"
+    
+    console.print(f"  Tests: {status_icon(entry.tests_passed)}")
+    console.print(f"  Lint: {status_icon(entry.lint_passed)}")
+    console.print(f"  Type Check: {status_icon(entry.type_check_passed)}")
+    
+    if entry.verification_notes:
+        console.print(f"\n[bold]Notes[/]")
+        console.print(f"  {entry.verification_notes}")
+    
+    # Hashes
+    console.print("\n[bold]Content Hashes[/]")
+    console.print(f"  Prompt: {entry.prompt_hash}")
+    console.print(f"  Context: {entry.context_hash or '-'}")

up/commands/review.py

@@ -0,0 +1,239 @@
+"""up review - AI-powered code review.
+
+Provides adversarial review of AI-generated code to catch:
+- Logic errors
+- Security issues
+- Performance problems
+- Best practice violations
+"""
+
+import subprocess
+import sys
+from pathlib import Path
+from typing import Optional
+
+import click
+from rich.console import Console
+from rich.panel import Panel
+from rich.markdown import Markdown
+
+from up.ai_cli import check_ai_cli, run_ai_prompt
+from up.core.checkpoint import get_checkpoint_manager
+
+console = Console()
+
+
+@click.command("review")
+@click.option("--checkpoint", "-c", help="Review changes since checkpoint")
+@click.option("--files", "-f", multiple=True, help="Specific files to review")
+@click.option("--focus", type=click.Choice(["security", "performance", "logic", "style", "all"]), 
+              default="all", help="Focus area for review")
+@click.option("--strict", is_flag=True, help="Strict mode - fail on any issue")
+@click.option("--model", "-m", help="AI model to use (claude, cursor)")
+def review_cmd(checkpoint: str, files: tuple, focus: str, strict: bool, model: str):
+    """AI-powered adversarial code review.
+    
+    Reviews recent AI-generated changes for potential issues.
+    Use after 'up start' to verify AI work before committing.
+    
+    \b
+    Examples:
+      up review                    # Review all changes
+      up review --checkpoint cp-1  # Review since checkpoint
+      up review --focus security   # Security-focused review
+      up review --files src/auth.py --strict
+    """
+    cwd = Path.cwd()
+    
+    # Get diff to review
+    if files:
+        # Review specific files
+        diff_content = _get_files_diff(cwd, list(files))
+        files_reviewed = list(files)
+    elif checkpoint:
+        # Review since checkpoint
+        diff_content = _get_checkpoint_diff(cwd, checkpoint)
+        files_reviewed = _get_changed_files(cwd, checkpoint)
+    else:
+        # Review all uncommitted changes
+        diff_content = _get_uncommitted_diff(cwd)
+        files_reviewed = _get_uncommitted_files(cwd)
+    
+    if not diff_content or diff_content.strip() == "":
+        console.print("[dim]No changes to review[/]")
+        return
+    
+    console.print(Panel.fit(
+        "[bold blue]AI Code Review[/]",
+        border_style="blue"
+    ))
+    
+    console.print(f"Files to review: [cyan]{len(files_reviewed)}[/]")
+    for f in files_reviewed[:5]:
+        console.print(f"  • {f}")
+    if len(files_reviewed) > 5:
+        console.print(f"  ... and {len(files_reviewed) - 5} more")
+    
+    # Check AI availability
+    cli_name = model
+    if not cli_name:
+        cli_name, available = check_ai_cli()
+        if not available:
+            console.print("\n[red]Error:[/] No AI CLI available")
+            console.print("Install Claude CLI or Cursor Agent for AI review")
+            return
+    
+    console.print(f"\n[yellow]Running review with {cli_name}...[/]")
+    
+    # Build review prompt
+    prompt = _build_review_prompt(diff_content, focus, strict)
+    
+    # Run AI review
+    result = run_ai_prompt(cwd, prompt, cli_name, timeout=180)
+    
+    if not result:
+        console.print("[red]Review failed - no response from AI[/]")
+        return
+    
+    # Display results
+    console.print("\n")
+    console.print(Panel(
+        Markdown(result),
+        title="Review Results",
+        border_style="yellow" if "issue" in result.lower() or "problem" in result.lower() else "green"
+    ))
+    
+    # Check for issues in strict mode
+    if strict:
+        issue_indicators = ["issue", "problem", "bug", "error", "vulnerability", "critical", "warning"]
+        has_issues = any(indicator in result.lower() for indicator in issue_indicators)
+        
+        if has_issues:
+            console.print("\n[red]⚠ Issues found in strict mode[/]")
+            console.print("Consider fixing issues before committing.")
+            sys.exit(1)
+        else:
+            console.print("\n[green]✓ No issues found[/]")
+
+
+def _get_uncommitted_diff(cwd: Path) -> str:
+    """Get diff of uncommitted changes."""
+    result = subprocess.run(
+        ["git", "diff", "HEAD"],
+        cwd=cwd,
+        capture_output=True,
+        text=True
+    )
+    return result.stdout if result.returncode == 0 else ""
+
+
+def _get_uncommitted_files(cwd: Path) -> list:
+    """Get list of uncommitted changed files."""
+    result = subprocess.run(
+        ["git", "diff", "--name-only", "HEAD"],
+        cwd=cwd,
+        capture_output=True,
+        text=True
+    )
+    if result.returncode == 0:
+        return [f for f in result.stdout.strip().split("\n") if f]
+    return []
+
+
+def _get_checkpoint_diff(cwd: Path, checkpoint: str) -> str:
+    """Get diff since checkpoint."""
+    manager = get_checkpoint_manager(cwd)
+    return manager.diff_from_checkpoint(checkpoint)
+
+
+def _get_changed_files(cwd: Path, checkpoint: str) -> list:
+    """Get files changed since checkpoint."""
+    tag_name = f"up-checkpoint/{checkpoint}"
+    result = subprocess.run(
+        ["git", "diff", "--name-only", tag_name, "HEAD"],
+        cwd=cwd,
+        capture_output=True,
+        text=True
+    )
+    if result.returncode == 0:
+        return [f for f in result.stdout.strip().split("\n") if f]
+    return []
+
+
+def _get_files_diff(cwd: Path, files: list) -> str:
+    """Get diff for specific files."""
+    result = subprocess.run(
+        ["git", "diff", "HEAD", "--"] + files,
+        cwd=cwd,
+        capture_output=True,
+        text=True
+    )
+    return result.stdout if result.returncode == 0 else ""
+
+
+def _build_review_prompt(diff: str, focus: str, strict: bool) -> str:
+    """Build the review prompt."""
+    # Truncate large diffs
+    max_chars = 15000
+    if len(diff) > max_chars:
+        diff = diff[:max_chars] + "\n\n[... diff truncated ...]"
+    
+    focus_instructions = {
+        "security": """Focus on security issues:
+- SQL injection, XSS, CSRF vulnerabilities
+- Hardcoded secrets or credentials
+- Insecure data handling
+- Authentication/authorization flaws
+- Input validation issues""",
+        
+        "performance": """Focus on performance issues:
+- O(n²) or worse algorithms
+- Memory leaks or inefficient memory use
+- Missing caching opportunities
+- Unnecessary database queries
+- Blocking operations in async code""",
+        
+        "logic": """Focus on logic issues:
+- Off-by-one errors
+- Race conditions
+- Null/undefined handling
+- Edge cases not covered
+- Incorrect state management""",
+        
+        "style": """Focus on code style:
+- Naming conventions
+- Function/class organization
+- Code duplication
+- Missing documentation
+- Type annotations""",
+        
+        "all": """Review comprehensively for:
+1. Security vulnerabilities
+2. Logic errors and bugs
+3. Performance issues
+4. Code style and best practices"""
+    }
+    
+    severity_note = ""
+    if strict:
+        severity_note = "\n\nIMPORTANT: This is a strict review. Flag ALL potential issues, even minor ones."
+    
+    return f"""You are a senior code reviewer conducting an adversarial review of AI-generated code.
+
+{focus_instructions.get(focus, focus_instructions["all"])}
+{severity_note}
+
+Review the following diff and provide:
+
+1. **Critical Issues** - Must fix before merging (security, data loss, crashes)
+2. **Warnings** - Should fix (bugs, performance, logic errors)  
+3. **Suggestions** - Nice to have (style, refactoring)
+4. **Summary** - Overall assessment (APPROVE / NEEDS CHANGES / REJECT)
+
+Be specific. Reference line numbers. Suggest fixes.
+
+```diff
+{diff}
+```
+
+Provide your review:"""