diff --git a/src/lysis/cli.py b/src/lysis/cli.py
new file mode 100644
index 0000000..cade0d2
--- /dev/null
+++ b/src/lysis/cli.py
@@ -0,0 +1,39 @@
+"""Rank a repo's open issues: python src\\lysis\\cli.py owner/repo [top]"""
+
+import sys
+
+import pandas as pd
+from dotenv import load_dotenv
+
+from lysis.features.rows import issue_row
+from lysis.ingestion.github_client import GitHubClient
+from lysis.models.evaluate import load_dataset
+from lysis.models.gbm import make_gbm_ranker
+
+
+def rank_repo(gh, full_name, ranker, top=10):
+    owner, name = full_name.split("/")
+    repo = gh.get(f"/repos/{full_name}")
+    issues = [i for i in gh.list_issues(owner, name, state="open") if not i.get("assignees")]
+    if not issues:
+        return pd.DataFrame(columns=["score", "number", "title", "url"])
+    df = pd.DataFrame([issue_row(i, repo) for i in issues])
+    df["score"] = list(ranker(df))
+    df = df.sort_values("score", ascending=False).head(top)
+    df["number"] = df["issue_number"]
+    df["title"] = df["title"].str[:60]
+    df["score"] = df["score"].round(3)
+    return df[["score", "number", "title", "url"]]
+
+
+def main():
+    load_dotenv()
+    if len(sys.argv) < 2:
+        sys.exit("usage: python src\\lysis\\cli.py owner/repo [top]")
+    top = int(sys.argv[2]) if len(sys.argv) > 2 else 10
+    ranker = make_gbm_ranker(load_dataset())
+    print(rank_repo(GitHubClient(), sys.argv[1], ranker, top).to_string(index=False))
+
+
+if __name__ == "__main__":
+    main()
\ No newline at end of file
diff --git a/src/lysis/features/rows.py b/src/lysis/features/rows.py
new file mode 100644
index 0000000..10d6cc2
--- /dev/null
+++ b/src/lysis/features/rows.py
@@ -0,0 +1,23 @@
+"""Build one model-input row from a GitHub issue and its repo."""
+
+
+def issue_row(issue, repo):
+    body = issue.get("body") or ""
+    return {
+        "repo": repo["full_name"],
+        "issue_number": issue["number"],
+        "title": issue.get("title") or "",
+        "url": issue.get("html_url", ""),
+        "labels": [x["name"].lower() for x in issue["labels"]],
+        "comments": issue["comments"],
+        "body_len": len(body),
+        "title_len": len(issue.get("title") or ""),
+        "has_code": int("```" in body),
+        "author_assoc": issue.get("author_association") or "NONE",
+        "stars": repo["stargazers_count"],
+        "forks": repo["forks_count"],
+        "repo_size": repo["size"],
+        "has_license": int(bool(repo.get("license"))),
+        "repo_created_at": repo["created_at"],
+        "created_at": issue["created_at"],
+    }
\ No newline at end of file
diff --git a/src/lysis/models/evaluate.py b/src/lysis/models/evaluate.py
index 5cd619b..b24fd05 100644
--- a/src/lysis/models/evaluate.py
+++ b/src/lysis/models/evaluate.py
@@ -8,6 +8,7 @@
 from sklearn.metrics import average_precision_score, roc_auc_score
 from sklearn.model_selection import GroupKFold
 
+from lysis.features.rows import issue_row
 from lysis.models.baseline import labels_only_ranker
 from lysis.models.gbm import make_gbm_ranker
 
@@ -22,26 +23,9 @@ def load_dataset(raw_dir="data/raw", labels_path="data/labels.csv"):
     for path in Path(raw_dir).glob("*.json"):
         d = json.loads(path.read_text())
         r = d["repo"]
-        name = r["full_name"]
         cutoff = min((p["created_at"] for p in d["merged_prs"]), default=None)
         rows += [
-        {
-                "repo": name,
-                "issue_number": i["number"],
-                "labels": [x["name"].lower() for x in i["labels"]],
-                "comments": i["comments"],
-                "body_len": len(i.get("body") or ""),
-                "title_len": len(i.get("title") or ""),
-                "text": f"{i.get('title') or ''} {i.get('body') or ''}"[:2000],
-                "has_code": int("```" in (i.get("body") or "")),
-                "author_assoc": i.get("author_association") or "NONE",
-                "stars": r["stargazers_count"],
-                "forks": r["forks_count"],
-                "repo_size": r["size"],
-                "has_license": int(bool(r.get("license"))),
-                "repo_created_at": r["created_at"],
-                "created_at": i["created_at"],
-            }
+            issue_row(i, r)
             for i in d["closed_issues"]
             if cutoff and i["created_at"] >= cutoff
         ]
@@ -49,7 +33,6 @@ def load_dataset(raw_dir="data/raw", labels_path="data/labels.csv"):
     df["first_timer"] = df["first_timer"].fillna(0).astype(int)
     return df
 
-
 def split_by_repo(df, test_frac=0.3, seed=42):
     repos = pd.Series(sorted(df["repo"].unique())).sample(frac=1, random_state=seed)
     test_repos = set(repos[: max(1, int(len(repos) * test_frac))])
diff --git a/tests/test_cli.py b/tests/test_cli.py
new file mode 100644
index 0000000..98712d9
--- /dev/null
+++ b/tests/test_cli.py
@@ -0,0 +1,33 @@
+from lysis.cli import rank_repo
+
+
+class FakeGH:
+    def get(self, path, params=None):
+        return {
+            "full_name": "a/b",
+            "stargazers_count": 10,
+            "forks_count": 1,
+            "size": 100,
+            "license": None,
+            "created_at": "2020-01-01T00:00:00Z",
+        }
+
+    def list_issues(self, owner, repo, state="open"):
+        base = {
+            "labels": [],
+            "comments": 0,
+            "body": "x",
+            "created_at": "2026-01-01T00:00:00Z",
+            "html_url": "u",
+            "assignees": [],
+        }
+        return [
+            {**base, "number": 1, "title": "one"},
+            {**base, "number": 2, "title": "two", "labels": [{"name": "good first issue"}]},
+            {**base, "number": 3, "title": "taken", "assignees": [{"login": "x"}]},
+        ]
+
+
+def test_rank_repo():
+    out = rank_repo(FakeGH(), "a/b", lambda df: df["labels"].apply(len), top=5)
+    assert list(out["number"]) == [2, 1]
\ No newline at end of file