feat(agents): migrate APEX to the Copilot SDK harness

.archive/pre-cpharness/README.md

@@ -0,0 +1,10 @@
+# Pre-Copilot-Harness Archive
+
+Historical VS Code Local prompt roots, split-harness adapters and superseded
+log procedures are retained here for reference, not discovery or execution.
+Their metadata, commands and relative links describe the previous layout.
+
+Active entry points are manual skills under `.github/skills/`; workflow routing
+uses named SDK workers with independent reviews, human gates and verified readiness.
+Do not add this archive to agent, skill, prompt, instruction or validator discovery.
+Existing project artifacts and recall state were not migrated or restamped.

.archive/pre-cpharness/parse-chat-logs.py

@@ -0,0 +1,273 @@
+#!/usr/bin/env python3
+"""Archived Local Copilot Chat log parser; not an active SDK entry point."""
+
+from __future__ import annotations
+
+import argparse
+import json
+import re
+import sys
+from dataclasses import asdict, dataclass, field
+from datetime import datetime
+from pathlib import Path
+
+CCREQ_PATTERN = re.compile(
+    r"(?P<timestamp>\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}\.\d{3})\s+"
+    r"\[(?P<level>\w+)\]\s+"
+    r"ccreq:(?P<request_id>[\w.]+)\s+\|\s+"
+    r"(?P<status>\w+)\s+\|\s+"
+    r"(?P<model>[^|]+?)\s+\|\s+"
+    r"(?P<latency_ms>\d+)ms\s+\|\s+"
+    r"\[(?P<request_type>[^\]]+)\]"
+)
+
+ERROR_PATTERN = re.compile(
+    r"(?P<timestamp>\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}\.\d{3})\s+"
+    r"\[error\]\s+(?P<message>.+)"
+)
+
+
+@dataclass
+class Request:
+    timestamp: str
+    request_id: str
+    status: str
+    model_requested: str
+    model_actual: str
+    latency_ms: int
+    request_type: str
+    session_dir: str
+
+
+@dataclass
+class ErrorEntry:
+    timestamp: str
+    message: str
+    session_dir: str
+
+
+@dataclass
+class SessionSummary:
+    session_dir: str
+    log_path: str
+    time_range: str
+    total_requests: int
+    total_errors: int
+    models: dict[str, int]
+    request_types: dict[str, int]
+    avg_latency_ms: float
+    max_latency_ms: int
+    p95_latency_ms: int
+    long_turns: int  # > 15s
+    burst_count: int  # gaps < 2s between consecutive requests
+    latency_trend: str  # escalating, stable, decreasing
+    requests: list[dict]
+    errors: list[dict]
+
+
+def parse_model_field(raw: str) -> tuple[str, str]:
+    """Split 'claude-opus-4.7 -> claude-opus-4-7' into (requested, actual)."""
+    raw = raw.strip()
+    if " -> " in raw:
+        parts = raw.split(" -> ", 1)
+        return parts[0].strip(), parts[1].strip()
+    return raw, raw
+
+
+def parse_log_file(log_path: Path, session_dir: str) -> tuple[list[Request], list[ErrorEntry]]:
+    """Extract request completions and errors from a single log file."""
+    requests: list[Request] = []
+    errors: list[ErrorEntry] = []
+
+    try:
+        text = log_path.read_text(encoding="utf-8", errors="replace")
+    except OSError:
+        return requests, errors
+
+    for line in text.splitlines():
+        m = CCREQ_PATTERN.search(line)
+        if m:
+            model_requested, model_actual = parse_model_field(m.group("model"))
+            requests.append(
+                Request(
+                    timestamp=m.group("timestamp"),
+                    request_id=m.group("request_id"),
+                    status=m.group("status"),
+                    model_requested=model_requested,
+                    model_actual=model_actual,
+                    latency_ms=int(m.group("latency_ms")),
+                    request_type=m.group("request_type"),
+                    session_dir=session_dir,
+                )
+            )
+            continue
+
+        em = ERROR_PATTERN.search(line)
+        if em:
+            errors.append(
+                ErrorEntry(
+                    timestamp=em.group("timestamp"),
+                    message=em.group("message")[:500],
+                    session_dir=session_dir,
+                )
+            )
+
+    return requests, errors
+
+
+def percentile(sorted_values: list[int], pct: float) -> int:
+    """Return the value at the given percentile from a sorted list."""
+    if not sorted_values:
+        return 0
+    idx = int(len(sorted_values) * pct / 100)
+    return sorted_values[min(idx, len(sorted_values) - 1)]
+
+
+def detect_latency_trend(requests: list[Request]) -> str:
+    """Determine if latency is escalating, stable, or decreasing over the session."""
+    agent_reqs = [r for r in requests if r.request_type == "panel/editAgent"]
+    if len(agent_reqs) < 4:
+        return "insufficient-data"
+
+    half = len(agent_reqs) // 2
+    first_half_avg = sum(r.latency_ms for r in agent_reqs[:half]) / half
+    second_half_avg = sum(r.latency_ms for r in agent_reqs[half:]) / (len(agent_reqs) - half)
+
+    ratio = second_half_avg / first_half_avg if first_half_avg > 0 else 1.0
+    if ratio > 1.3:
+        return "escalating"
+    if ratio < 0.7:
+        return "decreasing"
+    return "stable"
+
+
+def count_bursts(requests: list[Request]) -> int:
+    """Count rapid sequential requests (gap < 2 seconds)."""
+    bursts = 0
+    timestamps = []
+    for r in requests:
+        try:
+            ts = datetime.strptime(r.timestamp, "%Y-%m-%d %H:%M:%S.%f")
+            timestamps.append(ts)
+        except ValueError:
+            continue
+
+    for i in range(1, len(timestamps)):
+        gap = (timestamps[i] - timestamps[i - 1]).total_seconds()
+        if gap < 2.0:
+            bursts += 1
+    return bursts
+
+
+def summarize_session(
+    session_dir: str, log_path: str, requests: list[Request], errors: list[ErrorEntry]
+) -> SessionSummary:
+    """Build a session summary from parsed requests."""
+    models: dict[str, int] = {}
+    request_types: dict[str, int] = {}
+    latencies: list[int] = []
+
+    for r in requests:
+        models[r.model_actual] = models.get(r.model_actual, 0) + 1
+        request_types[r.request_type] = request_types.get(r.request_type, 0) + 1
+        latencies.append(r.latency_ms)
+
+    sorted_lat = sorted(latencies)
+    time_range = ""
+    if requests:
+        time_range = f"{requests[0].timestamp} — {requests[-1].timestamp}"
+
+    return SessionSummary(
+        session_dir=session_dir,
+        log_path=log_path,
+        time_range=time_range,
+        total_requests=len(requests),
+        total_errors=len(errors),
+        models=models,
+        request_types=request_types,
+        avg_latency_ms=round(sum(latencies) / len(latencies), 1) if latencies else 0,
+        max_latency_ms=max(latencies) if latencies else 0,
+        p95_latency_ms=percentile(sorted_lat, 95),
+        long_turns=sum(1 for lat in latencies if lat > 15000),
+        burst_count=count_bursts(requests),
+        latency_trend=detect_latency_trend(requests),
+        requests=[asdict(r) for r in requests],
+        errors=[asdict(e) for e in errors],
+    )
+
+
+def discover_log_files(log_dir: Path) -> list[tuple[Path, str]]:
+    """Find all Copilot Chat log files under the VS Code log directory."""
+    results = []
+    pattern = "**/GitHub.copilot-chat/GitHub Copilot Chat.log"
+    for log_path in sorted(log_dir.glob(pattern)):
+        session_dir = log_path.parent.parent.parent.name
+        results.append((log_path, session_dir))
+    return results
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser(
+        description="Parse Copilot Chat debug logs for context optimization analysis"
+    )
+    parser.add_argument(
+        "--log-dir",
+        type=Path,
+        default=Path.home() / ".vscode-server/data/logs",
+        help="VS Code log directory (default: ~/.vscode-server/data/logs)",
+    )
+    parser.add_argument(
+        "--output",
+        type=Path,
+        default=None,
+        help="Output JSON file path (default: stdout)",
+    )
+    parser.add_argument(
+        "--latest",
+        type=int,
+        default=0,
+        help="Only analyze the N most recent sessions (0 = all)",
+    )
+    parser.add_argument(
+        "--summary-only",
+        action="store_true",
+        help="Omit individual request details from output",
+    )
+    args = parser.parse_args()
+
+    log_files = discover_log_files(args.log_dir)
+    if not log_files:
+        print(f"No Copilot Chat logs found under {args.log_dir}", file=sys.stderr)
+        sys.exit(1)
+
+    if args.latest > 0:
+        log_files = log_files[-args.latest :]
+
+    sessions = []
+    for log_path, session_dir in log_files:
+        requests, errors = parse_log_file(log_path, session_dir)
+        if not requests:
+            continue
+        summary = summarize_session(session_dir, str(log_path), requests, errors)
+        if args.summary_only:
+            summary.requests = []
+            summary.errors = []
+        sessions.append(asdict(summary))
+
+    output = {
+        "generated": datetime.now().isoformat(),
+        "log_directory": str(args.log_dir),
+        "sessions_analyzed": len(sessions),
+        "sessions": sessions,
+    }
+
+    text = json.dumps(output, indent=2, default=str)
+    if args.output:
+        args.output.write_text(text, encoding="utf-8")
+        print(f"Wrote {len(sessions)} sessions to {args.output}", file=sys.stderr)
+    else:
+        print(text)
+
+
+if __name__ == "__main__":
+    main()