feed-the-machine 1.6.1 → 1.7.0

package/bin/migrate-eng-buddy-data.py

@@ -0,0 +1,875 @@
+#!/usr/bin/env python3
+"""Migrate eng-buddy markdown data to inbox.db ops tables.
+
+Reads ~/.claude/eng-buddy/{daily,patterns,capacity,stakeholders}/ markdown files,
+parses structured data with graceful fallback for inconsistent formatting, and
+inserts into the ops tracking tables in inbox.db.
+
+Usage:
+    python3 bin/migrate-eng-buddy-data.py --dry-run        # validate only
+    python3 bin/migrate-eng-buddy-data.py                  # full migration
+    python3 bin/migrate-eng-buddy-data.py --db-path /path/to/inbox.db
+"""
+
+import argparse
+import hashlib
+import json
+import os
+import re
+import sqlite3
+import subprocess
+import sys
+from datetime import datetime
+from pathlib import Path
+
+# ---------------------------------------------------------------------------
+# Constants
+# ---------------------------------------------------------------------------
+
+ENG_BUDDY_DIR = Path.home() / ".claude" / "eng-buddy"
+
+# Subdirs that get archived and parsed
+DATA_SUBDIRS = ["daily", "patterns", "capacity", "stakeholders"]
+
+# Severity keywords used to classify burnout / incident entries
+SEVERITY_KEYWORDS = {
+    "critical": ["🚨", "critical", "crisis", "emergency", "exhausted", "no sleep"],
+    "high": ["⚠️", "high", "warning", "stress", "blocked", "overload"],
+    "medium": ["medium", "moderate", "watch", "monitor"],
+    "low": ["low", "minor", "note"],
+}
+
+
+# ---------------------------------------------------------------------------
+# Helpers
+# ---------------------------------------------------------------------------
+
+
+def sha256(text: str) -> str:
+    return hashlib.sha256(text.encode()).hexdigest()
+
+
+def _first(patterns: list, text: str, group: int = 1, flags: int = re.IGNORECASE) -> str:
+    """Return the first match for any of the given regex patterns, or empty string."""
+    for pat in patterns:
+        m = re.search(pat, text, flags)
+        if m:
+            try:
+                return m.group(group).strip()
+            except IndexError:
+                return m.group(0).strip()
+    return ""
+
+
+def _all_matches(pattern: str, text: str, group: int = 1, flags: int = re.IGNORECASE) -> list:
+    return [m.group(group).strip() for m in re.finditer(pattern, text, flags)]
+
+
+def _infer_severity(text: str) -> str:
+    text_lower = text.lower()
+    for level in ("critical", "high", "medium", "low"):
+        for kw in SEVERITY_KEYWORDS[level]:
+            if kw.lower() in text_lower:
+                return level
+    return "medium"
+
+
+def _extract_date_from_filename(path: Path) -> str:
+    """Pull YYYY-MM-DD from filename if present."""
+    m = re.search(r"(\d{4}-\d{2}-\d{2})", path.stem)
+    return m.group(1) if m else ""
+
+
+def _split_sections(text: str, heading_re: str = r"^#{1,3} ") -> list:
+    """Split markdown into (heading, body) tuples."""
+    sections = []
+    current_heading = ""
+    current_lines = []
+    for line in text.splitlines():
+        if re.match(heading_re, line):
+            if current_heading or current_lines:
+                sections.append((current_heading, "\n".join(current_lines).strip()))
+            current_heading = line.lstrip("#").strip()
+            current_lines = []
+        else:
+            current_lines.append(line)
+    if current_heading or current_lines:
+        sections.append((current_heading, "\n".join(current_lines).strip()))
+    return sections
+
+
+def warn(msg: str) -> None:
+    print(f"  [WARN] {msg}", file=sys.stderr)
+
+
+# ---------------------------------------------------------------------------
+# Schema migration helpers
+# ---------------------------------------------------------------------------
+
+
+def _ensure_fingerprint_columns(conn: sqlite3.Connection) -> None:
+    """Add raw_content / content_sha256 columns if they don't exist yet."""
+    tables_needing_fingerprint = [
+        "capacity_logs",
+        "stakeholder_contacts",
+        "incidents",
+        "pattern_observations",
+        "follow_ups",
+        "burnout_indicators",
+    ]
+    cursor = conn.cursor()
+    for table in tables_needing_fingerprint:
+        # Check existing columns
+        cursor.execute(f"PRAGMA table_info({table})")
+        existing = {row[1] for row in cursor.fetchall()}
+        if "raw_content" not in existing:
+            cursor.execute(f"ALTER TABLE {table} ADD COLUMN raw_content TEXT")
+        if "content_sha256" not in existing:
+            cursor.execute(f"ALTER TABLE {table} ADD COLUMN content_sha256 TEXT")
+        if "source_file" not in existing and table not in ("pattern_observations",):
+            cursor.execute(f"ALTER TABLE {table} ADD COLUMN source_file TEXT")
+    conn.commit()
+
+
+# ---------------------------------------------------------------------------
+# Parsers
+# ---------------------------------------------------------------------------
+
+
+def parse_daily_logs(data_dir: Path, warnings: list) -> list:
+    """Parse ~/.claude/eng-buddy/daily/*.md → records for multiple tables."""
+    daily_dir = data_dir / "daily"
+    if not daily_dir.exists():
+        return []
+
+    records = {
+        "follow_ups": [],
+        "incidents": [],
+        "burnout_indicators": [],
+    }
+
+    for md_file in sorted(daily_dir.glob("*.md")):
+        text = md_file.read_text(errors="replace")
+        src = str(md_file)
+        date_str = _extract_date_from_filename(md_file) or _first(
+            [r"(\d{4}-\d{2}-\d{2})"], text
+        )
+
+        # Follow-ups: lines that look like tasks with "From X" or due-date markers
+        fu_pattern = r"[-*]\s+\*\*(?:From|By|Tomorrow|Next)\s+([^*]+?)\*\*[:\s]*(.+)"
+        for m in re.finditer(fu_pattern, text, re.IGNORECASE):
+            raw = m.group(0)
+            stakeholder = m.group(1).strip().rstrip(":")
+            topic = m.group(2).strip()
+            records["follow_ups"].append(
+                {
+                    "stakeholder": stakeholder,
+                    "topic": topic[:200],
+                    "due_date": date_str,
+                    "status": "pending",
+                    "notes": "",
+                    "raw_content": raw,
+                    "content_sha256": sha256(raw),
+                    "source_file": src,
+                }
+            )
+
+        # Incidents: blockers / bugs sections
+        blocker_section = re.search(
+            r"#{1,3}\s+(?:Blockers?|Bugs?|Issues?)[^\n]*\n(.*?)(?=\n#{1,3} |\Z)",
+            text,
+            re.DOTALL | re.IGNORECASE,
+        )
+        if blocker_section:
+            body = blocker_section.group(1)
+            # Each sub-heading inside that section = one incident
+            for sub_m in re.finditer(r"#{2,4}\s+(.+)\n(.*?)(?=\n#{2,4} |\Z)", body, re.DOTALL):
+                title = sub_m.group(1).strip()
+                details = sub_m.group(2).strip()
+                raw = sub_m.group(0)
+                if not title:
+                    continue
+                severity_str = _first(
+                    [r"\*\*Severity\*\*:\s*(\w+)", r"Severity:\s*(\w+)"], details
+                ) or _infer_severity(raw)
+                status = "open"
+                if re.search(r"(✅|RESOLVED|CLOSED|COMPLETE)", raw, re.IGNORECASE):
+                    status = "resolved"
+                records["incidents"].append(
+                    {
+                        "title": title[:200],
+                        "severity": severity_str.lower(),
+                        "status": status,
+                        "timeline": date_str,
+                        "root_cause": _first(
+                            [r"root[_ ]cause[:\s]+([^\n]+)", r"\*\*Root cause\*\*:\s*([^\n]+)"],
+                            details,
+                        )[:500],
+                        "resolution": _first(
+                            [r"resolution[:\s]+([^\n]+)", r"\*\*Resolution\*\*:\s*([^\n]+)"],
+                            details,
+                        )[:500],
+                        "raw_content": raw,
+                        "content_sha256": sha256(raw),
+                        "source_file": src,
+                    }
+                )
+
+        # Burnout indicators from "Red Flags" / "Burnout" sections
+        burnout_section = re.search(
+            r"#{1,3}\s+(?:Red Flags?|Burnout)[^\n]*\n(.*?)(?=\n#{1,3} |\Z)",
+            text,
+            re.DOTALL | re.IGNORECASE,
+        )
+        if burnout_section:
+            for line in burnout_section.group(1).splitlines():
+                line = line.strip()
+                if not line or not re.match(r"[-*🚨⚠️]", line):
+                    continue
+                raw = line
+                severity = _infer_severity(line)
+                indicator = re.sub(r"^[-*🚨⚠️]+\s*", "", line).strip()
+                records["burnout_indicators"].append(
+                    {
+                        "date": date_str,
+                        "indicator": indicator[:200],
+                        "severity": severity,
+                        "details": "",
+                        "raw_content": raw,
+                        "content_sha256": sha256(raw),
+                        "source_file": src,
+                    }
+                )
+
+    return records
+
+
+def parse_patterns(data_dir: Path, warnings: list) -> list:
+    """Parse ~/.claude/eng-buddy/patterns/*.md → pattern_observations rows."""
+    patterns_dir = data_dir / "patterns"
+    if not patterns_dir.exists():
+        return []
+
+    rows = []
+    for md_file in sorted(patterns_dir.glob("*.md")):
+        text = md_file.read_text(errors="replace")
+        src = str(md_file)
+
+        # Infer type from filename
+        fname = md_file.stem.lower()
+        if "success" in fname:
+            pat_type = "success"
+        elif "failure" in fname or "anti" in fname:
+            pat_type = "failure"
+        elif "burnout" in fname:
+            pat_type = "burnout"
+        elif "recurring" in fname:
+            pat_type = "recurring"
+        elif "task" in fname or "execution" in fname:
+            pat_type = "execution"
+        elif "time" in fname:
+            pat_type = "time_estimate"
+        else:
+            pat_type = "observation"
+
+        # Each ### heading = one pattern entry
+        for heading, body in _split_sections(text, r"^#{2,4} "):
+            if not heading or not body:
+                continue
+            raw = f"### {heading}\n{body}"
+            date_str = _first(
+                [r"(\d{4}-\d{2}-\d{2})", r"\((\d{4}-\d{2}-\d{2})\)"], heading + " " + body
+            )
+            # Frequency from explicit field or occurrence count
+            freq_raw = _first(
+                [r"count[:\s]+(\d+)", r"frequency[:\s]+(\d+)", r"occurrences?[:\s]+(\d+)"],
+                body,
+            )
+            frequency = int(freq_raw) if freq_raw.isdigit() else 1
+
+            description = body[:1000]
+            evidence = _first(
+                [r"(?:evidence|example|data)[:\s]+(.+?)(?:\n\n|\Z)", r"\*\*Result\*\*:\s*(.+)"],
+                body,
+                flags=re.DOTALL | re.IGNORECASE,
+            )[:500]
+
+            rows.append(
+                {
+                    "type": pat_type,
+                    "title": heading[:200],
+                    "description": description,
+                    "confidence": None,
+                    "evidence": evidence,
+                    "frequency": frequency,
+                    "first_seen": date_str,
+                    "last_seen": date_str,
+                    "source_file": src,
+                    "raw_content": raw,
+                    "content_sha256": sha256(raw),
+                }
+            )
+
+    return rows
+
+
+def parse_capacity(data_dir: Path, warnings: list) -> list:
+    """Parse ~/.claude/eng-buddy/capacity/*.md → capacity_logs rows.
+
+    Skips burnout-indicators.md (handled separately by parse_burnout_indicators).
+    """
+    capacity_dir = data_dir / "capacity"
+    if not capacity_dir.exists():
+        return []
+
+    SKIP_FILES = {"burnout-indicators.md"}
+    rows = []
+    any_file_had_rows = False
+
+    for md_file in sorted(capacity_dir.glob("*.md")):
+        if md_file.name in SKIP_FILES:
+            continue
+
+        text = md_file.read_text(errors="replace")
+        src = str(md_file)
+        date_ctx = _extract_date_from_filename(md_file)
+        file_rows = []
+
+        # Pattern 1: "- **Total capacity**: 40 hours"
+        for m in re.finditer(
+            r"-?\s*\*\*([^*]+?)\*\*[:\s]+([\d.]+)\s*(?:hours?)?", text, re.IGNORECASE
+        ):
+            metric = m.group(1).strip()
+            try:
+                value = float(m.group(2))
+            except ValueError:
+                warnings.append(f"{md_file.name}: cannot parse value '{m.group(2)}' for '{metric}'")
+                continue
+            raw = m.group(0)
+            date_str = date_ctx or _first([r"(\d{4}-\d{2}-\d{2})"], text)
+            file_rows.append(
+                {
+                    "date": date_str,
+                    "metric": metric[:100],
+                    "value": value,
+                    "notes": "",
+                    "raw_content": raw,
+                    "content_sha256": sha256(raw),
+                    "source_file": src,
+                }
+            )
+
+        # Pattern 2: "sleep: 4 hours" / "Sleep (4h)" / "4-hour sleep"
+        for m in re.finditer(
+            r"(\bsleep\b[^:,\n]*)[:\s]+([\d.]+)\s*(?:hours?|h\b)"
+            r"|(\d+\.?\d*)[- ]hour(?:s)?\s+sleep",
+            text,
+            re.IGNORECASE,
+        ):
+            metric = "sleep_hours"
+            raw = m.group(0)
+            # Extract numeric value from whichever capture group matched
+            val_str = m.group(2) if m.group(2) else m.group(3)
+            try:
+                value = float(val_str)
+            except (ValueError, TypeError):
+                continue
+            date_str = date_ctx or _first([r"(\d{4}-\d{2}-\d{2})"], raw + "\n" + text[:500])
+            note = (m.group(1) or "").strip()
+            file_rows.append(
+                {
+                    "date": date_str,
+                    "metric": metric,
+                    "value": value,
+                    "notes": note[:200],
+                    "raw_content": raw,
+                    "content_sha256": sha256(raw),
+                    "source_file": src,
+                }
+            )
+
+        if not file_rows:
+            warnings.append(f"{md_file.name}: no capacity metrics extracted")
+        else:
+            any_file_had_rows = True
+
+        rows.extend(file_rows)
+
+    return rows
+
+
+def parse_burnout_indicators(data_dir: Path, warnings: list) -> list:
+    """Parse capacity/burnout-indicators.md → burnout_indicators rows."""
+    bi_file = data_dir / "capacity" / "burnout-indicators.md"
+    if not bi_file.exists():
+        return []
+
+    text = bi_file.read_text(errors="replace")
+    src = str(bi_file)
+    rows = []
+
+    # Each ### subheading is an indicator category; each bullet is an entry
+    for heading, body in _split_sections(text, r"^#{2,4} "):
+        if not heading or not body:
+            continue
+        if re.search(r"(action|recommendation|immediate|recovery)", heading, re.IGNORECASE):
+            continue  # skip action sections
+
+        for line in body.splitlines():
+            line = line.strip()
+            if not line or not re.match(r"[-*]|🚨|⚠️|✅", line):
+                continue
+            # Skip sub-bullet context lines (indented continuation)
+            raw = line
+            severity = _infer_severity(heading + " " + line)
+            indicator = re.sub(r"^[-*🚨⚠️✅]+\s*", "", line).strip()
+            if not indicator:
+                continue
+
+            # Date: from line or from heading
+            date_str = _first([r"(\d{4}-\d{2}-\d{2})"], heading + " " + line)
+
+            rows.append(
+                {
+                    "date": date_str,
+                    "indicator": indicator[:200],
+                    "severity": severity,
+                    "details": heading[:200],
+                    "raw_content": raw,
+                    "content_sha256": sha256(raw),
+                    "source_file": src,
+                }
+            )
+
+    return rows
+
+
+def parse_incidents_dir(data_dir: Path, warnings: list) -> list:
+    """Parse ~/.claude/eng-buddy/incidents/*.md → incidents rows."""
+    incidents_dir = data_dir / "incidents"
+    if not incidents_dir.exists():
+        return []
+
+    SKIP_FILES = {"incident-index.md"}
+    rows = []
+
+    for md_file in sorted(incidents_dir.glob("*.md")):
+        if md_file.name in SKIP_FILES:
+            continue
+
+        text = md_file.read_text(errors="replace")
+        src = str(md_file)
+        raw_full = text[:2000]  # fingerprint first 2KB
+
+        # Title: first H1 or from filename
+        title = _first([r"^#\s+(?:Incident[:\s]*)?(.+)$"], text, flags=re.MULTILINE) or md_file.stem
+
+        # Remove emoji/status suffix from title
+        title = re.sub(r"\s*[✅❌🚨⚠️]+.*$", "", title).strip()
+
+        # Date
+        date_str = _first([r"\*\*Date\*\*:\s*(\d{4}-\d{2}-\d{2})", r"(\d{4}-\d{2}-\d{2})"], text)
+
+        # Severity
+        severity_raw = _first(
+            [r"\*\*Severity\*\*:\s*([^\n]+)", r"severity[:\s]+([^\n,]+)"], text
+        )
+        severity = severity_raw.lower().split()[0] if severity_raw else _infer_severity(text[:500])
+        # Normalize e.g. "Critical → Resolved" → "critical"
+        severity = re.sub(r"[^a-z].*", "", severity)
+
+        # Status
+        status = "open"
+        if re.search(r"(✅\s*COMPLETE|RESOLVED|CLOSED|status.*:\s*✅)", text, re.IGNORECASE):
+            status = "resolved"
+
+        # Timeline: the explicit Timeline section
+        timeline = _first(
+            [r"## Timeline\n(.*?)(?=\n## |\Z)"], text, flags=re.DOTALL | re.IGNORECASE
+        )[:500]
+
+        # Root cause
+        root_cause = _first(
+            [r"Root Cause[:\s]+([^\n]+)", r"\*\*Root Cause\*\*:\s*([^\n]+)"], text
+        )[:300]
+
+        # Resolution
+        resolution = _first(
+            [r"Resolution[:\s]+([^\n]+)", r"\*\*Resolution\*\*:\s*([^\n]+)"], text
+        )[:300]
+
+        rows.append(
+            {
+                "title": title[:200],
+                "severity": severity[:50],
+                "status": status,
+                "timeline": timeline or date_str,
+                "root_cause": root_cause,
+                "resolution": resolution,
+                "raw_content": raw_full,
+                "content_sha256": sha256(raw_full),
+                "source_file": src,
+            }
+        )
+
+    return rows
+
+
+def parse_stakeholders(data_dir: Path, warnings: list) -> list:
+    """Parse ~/.clone/eng-buddy/stakeholders/*.md → stakeholder_contacts rows."""
+    stakeholders_dir = data_dir / "stakeholders"
+    if not stakeholders_dir.exists():
+        return []
+
+    rows = []
+    for md_file in sorted(stakeholders_dir.glob("*.md")):
+        text = md_file.read_text(errors="replace")
+        src = str(md_file)
+
+        # Sections delineated by ### Name headings
+        for heading, body in _split_sections(text, r"^#{2,3} "):
+            if not heading or not body:
+                continue
+            raw = f"### {heading}\n{body}"
+
+            # Name
+            name = heading.strip()
+            if len(name) > 100 or re.search(r"(pending|waiting|vendor|overview|notes)", name, re.IGNORECASE):
+                continue  # skip section headings that aren't person names
+
+            # Role
+            role = _first(
+                [
+                    r"\*\*Role\*\*:\s*([^\n]+)",
+                    r"Role:\s*([^\n]+)",
+                    r"\*\*Title\*\*:\s*([^\n]+)",
+                ],
+                body,
+            )[:200]
+
+            # Preferences / communication style
+            preferences = _first(
+                [
+                    r"\*\*Communication preference\*\*:\s*([^\n]+)",
+                    r"Communication[^:]*:\s*([^\n]+)",
+                    r"\*\*Prefer[^*]*\*\*:\s*([^\n]+)",
+                ],
+                body,
+            )[:300]
+
+            # Last contact date
+            last_contact = _first(
+                [
+                    r"\*\*Last contact\*\*:\s*(\d{4}-\d{2}-\d{2})",
+                    r"(\d{4}-\d{2}-\d{2})",
+                ],
+                body,
+            )
+
+            rows.append(
+                {
+                    "name": name[:100],
+                    "role": role,
+                    "preferences": preferences,
+                    "last_contact": last_contact,
+                    "raw_content": raw,
+                    "content_sha256": sha256(raw),
+                    "source_file": src,
+                }
+            )
+
+    return rows
+
+
+# ---------------------------------------------------------------------------
+# Archive
+# ---------------------------------------------------------------------------
+
+
+def archive_originals(data_dir: Path) -> Path:
+    """Tar the data subdirs that exist and return the archive path."""
+    today = datetime.now().strftime("%Y-%m-%d")
+    archive_path = data_dir / f"archive-{today}.tar.gz"
+
+    dirs_to_archive = [str(data_dir / d) for d in DATA_SUBDIRS if (data_dir / d).exists()]
+    if not dirs_to_archive:
+        print("  [archive] No source dirs found — skipping archive step.")
+        return archive_path
+
+    cmd = ["tar", "czf", str(archive_path)] + dirs_to_archive
+    print(f"  [archive] {' '.join(cmd)}")
+    result = subprocess.run(cmd, capture_output=True, text=True)
+    if result.returncode != 0:
+        print(f"  [archive] WARNING: tar failed: {result.stderr}", file=sys.stderr)
+    else:
+        print(f"  [archive] Created {archive_path}")
+    return archive_path
+
+
+# ---------------------------------------------------------------------------
+# DB insertion
+# ---------------------------------------------------------------------------
+
+
+def _insert_rows(conn, table: str, rows: list, dry_run: bool) -> int:
+    if not rows:
+        return 0
+    # Build column list from first row keys (exclude None-value-only columns)
+    cols = list(rows[0].keys())
+    placeholders = ", ".join(["?"] * len(cols))
+    col_str = ", ".join(cols)
+    sql = f"INSERT OR IGNORE INTO {table} ({col_str}) VALUES ({placeholders})"
+
+    if dry_run:
+        return len(rows)
+
+    cursor = conn.cursor()
+    for row in rows:
+        values = [row.get(c) for c in cols]
+        try:
+            cursor.execute(sql, values)
+        except sqlite3.OperationalError as e:
+            warn(f"Insert into {table} failed: {e} — row keys: {list(row.keys())}")
+    return len(rows)
+
+
+# ---------------------------------------------------------------------------
+# Reconciliation
+# ---------------------------------------------------------------------------
+
+
+def reconcile(conn, table: str, expected_sources: set) -> bool:
+    """Verify that all expected source files appear in the DB."""
+    cursor = conn.execute(f"SELECT DISTINCT source_file FROM {table}")
+    db_sources = {row[0] for row in cursor.fetchall() if row[0]}
+    missing = expected_sources - db_sources
+    if missing:
+        for f in sorted(missing):
+            warn(f"Reconciliation: {table} missing source_file '{f}'")
+        return False
+    return True
+
+
+# ---------------------------------------------------------------------------
+# Dry-run report
+# ---------------------------------------------------------------------------
+
+
+def print_report(all_data: dict, warnings: list) -> None:
+    print("\n" + "=" * 60)
+    print("DRY-RUN VALIDATION REPORT")
+    print("=" * 60)
+
+    total = 0
+    for table, rows in all_data.items():
+        if isinstance(rows, list):
+            count = len(rows)
+        elif isinstance(rows, dict):
+            count = sum(len(v) for v in rows.values())
+        else:
+            count = 0
+
+        sources = set()
+        if isinstance(rows, list):
+            sources = {r.get("source_file", "") for r in rows}
+        elif isinstance(rows, dict):
+            for v in rows.values():
+                sources |= {r.get("source_file", "") for r in v}
+
+        sha_coverage = 0
+        flat_rows = rows if isinstance(rows, list) else [r for v in rows.values() for r in v]
+        sha_coverage = sum(1 for r in flat_rows if r.get("content_sha256"))
+        field_pct = (sha_coverage / count * 100) if count else 0
+
+        print(f"\n  {table}:")
+        print(f"    rows         : {count}")
+        print(f"    source files : {len(sources)}")
+        print(f"    sha256 cover : {sha_coverage}/{count} ({field_pct:.0f}%)")
+        total += count
+
+    print(f"\n  TOTAL ROWS: {total}")
+
+    if warnings:
+        print(f"\n  WARNINGS ({len(warnings)}):")
+        for w in warnings:
+            print(f"    - {w}")
+    else:
+        print("\n  No warnings.")
+
+    print("=" * 60)
+
+
+# ---------------------------------------------------------------------------
+# Main
+# ---------------------------------------------------------------------------
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser(
+        description="Migrate eng-buddy markdown to inbox.db ops tables."
+    )
+    parser.add_argument(
+        "--dry-run",
+        action="store_true",
+        help="Parse and validate without writing to the DB.",
+    )
+    parser.add_argument(
+        "--db-path",
+        help="Path to inbox.db (auto-detected from brain.py path if not provided).",
+    )
+    parser.add_argument(
+        "--data-dir",
+        default=str(ENG_BUDDY_DIR),
+        help="Path to eng-buddy data directory (default: ~/.claude/eng-buddy).",
+    )
+    args = parser.parse_args()
+
+    data_dir = Path(args.data_dir).expanduser()
+
+    # Resolve DB path
+    if args.db_path:
+        db_path = Path(args.db_path).expanduser()
+    else:
+        db_path = Path.home() / ".claude" / "eng-buddy" / "inbox.db"
+
+    if not db_path.exists():
+        print(f"ERROR: inbox.db not found at {db_path}", file=sys.stderr)
+        print("Pass --db-path to specify location.", file=sys.stderr)
+        return 1
+
+    print(f"eng-buddy migration")
+    print(f"  data_dir : {data_dir}")
+    print(f"  db_path  : {db_path}")
+    print(f"  dry_run  : {args.dry_run}")
+    print()
+
+    # ---- Parse ----
+    warnings: list = []
+    print("Parsing daily logs...")
+    daily_records = parse_daily_logs(data_dir, warnings)
+
+    print("Parsing pattern files...")
+    pattern_rows = parse_patterns(data_dir, warnings)
+
+    print("Parsing capacity files...")
+    capacity_rows = parse_capacity(data_dir, warnings)
+
+    print("Parsing burnout indicators...")
+    burnout_rows = parse_burnout_indicators(data_dir, warnings)
+
+    print("Parsing incidents directory...")
+    incident_rows = parse_incidents_dir(data_dir, warnings)
+
+    print("Parsing stakeholder files...")
+    stakeholder_rows = parse_stakeholders(data_dir, warnings)
+
+    # daily_records is a dict of lists keyed by table name
+    follow_up_rows = daily_records.get("follow_ups", [])
+    # Merge incidents from daily logs + incidents dir (daily has inline blockers)
+    incident_rows = incident_rows + daily_records.get("incidents", [])
+    # Merge burnout from capacity/burnout-indicators.md + inline daily log sections
+    burnout_rows = burnout_rows + daily_records.get("burnout_indicators", [])
+
+    all_data = {
+        "capacity_logs": capacity_rows,
+        "stakeholder_contacts": stakeholder_rows,
+        "incidents": incident_rows,
+        "pattern_observations": pattern_rows,
+        "follow_ups": follow_up_rows,
+        "burnout_indicators": burnout_rows,
+    }
+
+    if args.dry_run:
+        print_report(all_data, warnings)
+        return 0
+
+    # ---- Archive originals ----
+    print("\nArchiving originals...")
+    archive_originals(data_dir)
+
+    # ---- Write to DB (single transaction) ----
+    print("\nWriting to DB...")
+    conn = sqlite3.connect(str(db_path))
+    try:
+        # Ensure fingerprint columns exist
+        _ensure_fingerprint_columns(conn)
+
+        # Begin single transaction
+        conn.execute("BEGIN")
+
+        counts = {}
+        counts["capacity_logs"] = _insert_rows(conn, "capacity_logs", capacity_rows, False)
+        counts["stakeholder_contacts"] = _insert_rows(
+            conn, "stakeholder_contacts", stakeholder_rows, False
+        )
+        counts["incidents"] = _insert_rows(conn, "incidents", incident_rows, False)
+        counts["pattern_observations"] = _insert_rows(
+            conn, "pattern_observations", pattern_rows, False
+        )
+        counts["follow_ups"] = _insert_rows(conn, "follow_ups", follow_up_rows, False)
+        counts["burnout_indicators"] = _insert_rows(
+            conn, "burnout_indicators", burnout_rows, False
+        )
+
+        conn.commit()
+        print("  Transaction committed.")
+
+    except Exception as e:
+        conn.rollback()
+        print(f"ERROR: Transaction rolled back: {e}", file=sys.stderr)
+        conn.close()
+        return 2
+
+    # ---- Reconciliation pass ----
+    print("\nRunning reconciliation pass...")
+    ok = True
+
+    table_source_map = {
+        "capacity_logs": {r["source_file"] for r in capacity_rows if r.get("source_file")},
+        "stakeholder_contacts": {
+            r["source_file"] for r in stakeholder_rows if r.get("source_file")
+        },
+        "incidents": {r["source_file"] for r in incident_rows if r.get("source_file")},
+        "pattern_observations": {
+            r["source_file"] for r in pattern_rows if r.get("source_file")
+        },
+        "follow_ups": {r["source_file"] for r in follow_up_rows if r.get("source_file")},
+        "burnout_indicators": {
+            r["source_file"] for r in burnout_rows if r.get("source_file")
+        },
+    }
+
+    for table, expected_sources in table_source_map.items():
+        if not expected_sources:
+            continue
+        passed = reconcile(conn, table, expected_sources)
+        status = "OK" if passed else "MISMATCH"
+        print(f"  {table}: {status} ({len(expected_sources)} source files)")
+        if not passed:
+            ok = False
+
+    conn.close()
+
+    # ---- Summary ----
+    print("\nMigration summary:")
+    total = 0
+    for table, n in counts.items():
+        print(f"  {table}: {n} rows inserted")
+        total += n
+    print(f"  TOTAL: {total} rows")
+
+    if warnings:
+        print(f"\nWarnings ({len(warnings)}):")
+        for w in warnings:
+            print(f"  - {w}")
+
+    if not ok:
+        print("\nERROR: Reconciliation mismatch — check warnings above.", file=sys.stderr)
+        return 3
+
+    print("\nMigration complete.")
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())