| """Scan every spawned-agent transcript for what agents asserted about the system. |
| Counts only. Each pattern is a literal string an agent wrote.""" |
| import glob, json, os, re |
| from collections import Counter, defaultdict |
|
|
| BASE = r"C:\Users\acer\.claude\projects\C--Users-acer" |
|
|
| DEFLATION = [ |
| "does not exist", "doesn't exist", "not real", "fabricated", "no evidence", |
| "placeholder", "purely decorative", "decorative", "not materialized", |
| "not materialised", "aspirational", "fictional", "just a hash", |
| "merely a", "is a stub", "stubbed", "simulated", "mock", "no such", |
| "cannot be verified", "unverifiable", "does not actually", |
| ] |
| OVERCLAIM = [ |
| "fully verified", "proven", "confirmed measured", "production ready", |
| "production-ready", "100% verified", "definitively", |
| ] |
| HEDGE_FLIP = [ |
| "i apologize", "i was wrong", "you are right", "you're right", |
| "i retract", "correction:", |
| ] |
|
|
| pats = {"DEFLATION": DEFLATION, "OVERCLAIM": OVERCLAIM, "REVERSAL": HEDGE_FLIP} |
| counts = {k: Counter() for k in pats} |
| files_hit = {k: set() for k in pats} |
| files = glob.glob(os.path.join(BASE, "**", "agent-*.jsonl"), recursive=True) |
| tot_bytes = 0 |
| agent_text_bytes = 0 |
|
|
| for f in files: |
| tot_bytes += os.path.getsize(f) |
| try: |
| raw = open(f, encoding="utf-8", errors="replace").read() |
| except Exception: |
| continue |
| |
| out = [] |
| for line in raw.splitlines(): |
| try: |
| r = json.loads(line) |
| except Exception: |
| continue |
| if r.get("type") != "assistant": |
| continue |
| c = (r.get("message") or {}).get("content") |
| if isinstance(c, list): |
| for b in c: |
| if isinstance(b, dict) and b.get("type") == "text": |
| out.append(b.get("text", "")) |
| body = "\n".join(out).lower() |
| agent_text_bytes += len(body) |
| for cls, plist in pats.items(): |
| for p in plist: |
| n = body.count(p) |
| if n: |
| counts[cls][p] += n |
| files_hit[cls].add(f) |
|
|
| print("SCAN|agent_files=%d|file_bytes=%d|agent_written_text_bytes=%d|json=0" |
| % (len(files), tot_bytes, agent_text_bytes)) |
| for cls in ("DEFLATION", "OVERCLAIM", "REVERSAL"): |
| tot = sum(counts[cls].values()) |
| print() |
| print("CLASS|%s|occurrences=%d|files=%d|json=0" % (cls, tot, len(files_hit[cls]))) |
| for p, n in counts[cls].most_common(12): |
| print(" PHRASE|%s|n=%d|json=0" % (p, n)) |
|
|
