Present completion research as web document

tools/check_project.py

@@ -4,6 +4,7 @@
 from __future__ import annotations
 
 import ast
+import hashlib
 import json
 from collections import Counter
 from pathlib import Path
@@ -21,12 +22,19 @@ REQUIRED_FILES = (
     "data/summary.json",
     "research/memory/findings.md",
     "research/memory/evidence-ledger.md",
+    "research/completion-verification/findings.md",
+    "research/completion-verification/evidence-ledger.md",
     "research/evaluation/k1412-agent-evaluation-v1.md",
     "data/evaluation/k1412-agent-eval-task-contracts-v1.json",
     "tools/evaluation/score_agent_runs.py",
     "evaluation-site/index.html",
+    "evaluation-site/evaluation-v1.html",
     "evaluation-site/assets/styles.css",
     "evaluation-site/assets/app.js",
+    "evaluation-site/assets/research.css",
+    "evaluation-site/assets/research.js",
+    "evaluation-site/assets/evidence-data.js",
+    "evaluation-site/build_research_data.py",
     "evaluation-site/Dockerfile",
     "evaluation-site/nginx.conf",
     "web/app.py",
@@ -96,7 +104,10 @@ def check_research_data() -> dict[str, int]:
 
 
 def check_python_syntax() -> int:
-    paths = sorted((ROOT / "tools").rglob("*.py")) + [ROOT / "web" / "app.py"]
+    paths = sorted((ROOT / "tools").rglob("*.py")) + [
+        ROOT / "web" / "app.py",
+        ROOT / "evaluation-site" / "build_research_data.py",
+    ]
     for path in paths:
         ast.parse(path.read_text(encoding="utf-8"), filename=str(path))
     return len(paths)
@@ -143,16 +154,52 @@ def check_evaluation_contract() -> int:
     return len(tasks)
 
 
-def check_evaluation_site() -> None:
+def check_evaluation_site() -> int:
     html = (ROOT / "evaluation-site/index.html").read_text(encoding="utf-8")
+    archived_html = (ROOT / "evaluation-site/evaluation-v1.html").read_text(encoding="utf-8")
     javascript = (ROOT / "evaluation-site/assets/app.js").read_text(encoding="utf-8")
+    research_javascript = (ROOT / "evaluation-site/assets/research.js").read_text(encoding="utf-8")
+    evidence_javascript = (ROOT / "evaluation-site/assets/evidence-data.js").read_text(
+        encoding="utf-8"
+    )
+    ledger = (ROOT / "research/completion-verification/evidence-ledger.md").read_text(
+        encoding="utf-8"
+    )
     dockerfile = (ROOT / "evaluation-site/Dockerfile").read_text(encoding="utf-8")
     nginx = (ROOT / "evaluation-site/nginx.conf").read_text(encoding="utf-8")
-    require("不是真实跑分" in html, "evaluation simulation must keep its evidence boundary visible")
-    require("尚未运行真实模型对比" in html, "evaluation report overstates current evidence")
+    require("Agent 到底凭什么说" in html, "completion research document is missing")
+    require("自有跑分" in html and ">0<" in html, "research page overstates benchmark evidence")
+    require("completion certificate" in html.lower(), "research synthesis boundary is missing")
+    require(
+        "不是真实跑分" in archived_html,
+        "archived evaluation simulation must keep its evidence boundary visible",
+    )
+    require(
+        "尚未运行真实模型对比" in archived_html,
+        "archived evaluation report overstates current evidence",
+    )
     require("const TASKS" in javascript, "evaluation task browser is missing")
+    require("K1412_EVIDENCE" in research_javascript, "research evidence browser is missing")
+
+    prefix = "window.K1412_EVIDENCE="
+    require(evidence_javascript.startswith("/* Generated") and prefix in evidence_javascript, "bad evidence data")
+    payload_text = evidence_javascript.split(prefix, 1)[1].strip()
+    require(payload_text.endswith(";"), "bad evidence data terminator")
+    payload = json.loads(payload_text[:-1])
+    counts = payload.get("counts") or {}
+    records = payload.get("records") or []
+    require(counts.get("pool") == len(records) == 40, "completion evidence pool count mismatch")
+    require(counts.get("audited") == 37, "completion audited count mismatch")
+    require(counts.get("core") == 26, "completion core count mismatch")
+    require(
+        payload.get("source_sha256") == hashlib.sha256(ledger.encode("utf-8")).hexdigest(),
+        "completion evidence browser data is stale",
+    )
+
     require("install -d -m 0755" in dockerfile, "static asset directory must remain traversable")
+    require("evaluation-v1.html" in dockerfile, "archived evaluation page is not packaged")
     require('location = /health' in nginx, "evaluation site health endpoint is missing")
+    return int(counts["audited"])
 
 
 def main() -> int:
@@ -163,7 +210,7 @@ def main() -> int:
     check_ollama_contract()
     check_experiment_state()
     evaluation_tasks = check_evaluation_contract()
-    check_evaluation_site()
+    completion_evidence = check_evaluation_site()
     print(
         json.dumps(
             {
@@ -171,6 +218,7 @@ def main() -> int:
                 "collections": collections,
                 "research": research,
                 "evaluation_tasks": evaluation_tasks,
+                "completion_evidence": completion_evidence,
                 "evaluation_site": True,
                 "python_files_checked": python_files,
             },