From f0deca71de6659da1b4dfb211a88c092943a9bca Mon Sep 17 00:00:00 2001
From: Cid-oe <cid066a86@gmail.com>
Date: Wed, 9 Sep 2026 23:07:18 +0530
Subject: [PATCH] fix(trace): exclude non-HTTP URL schemes like data: URLs from
 HTTP accesses

---
 packages/trace/src/index.ts | 16 ++++++++++++----
 tests/trace.test.mjs        | 37 +++++++++++++++++++++++++++++++++++++
 2 files changed, 49 insertions(+), 4 deletions(-)

diff --git a/packages/trace/src/index.ts b/packages/trace/src/index.ts
index 59b0d32..bd8bb2e 100644
--- a/packages/trace/src/index.ts
+++ b/packages/trace/src/index.ts
@@ -160,10 +160,18 @@ function recordFileAccess(access: "read" | "write", selector: fs.PathOrFileDescr
 }
 
 function fetchSelector(input: Parameters<typeof fetch>[0]): string | undefined {
-  if (typeof input === "string") return input;
-  if (input instanceof URL) return input.toString();
-  if (typeof Request !== "undefined" && input instanceof Request) return input.url;
-  return undefined;
+  let raw: string | undefined;
+  if (typeof input === "string") raw = input;
+  else if (input instanceof URL) raw = input.toString();
+  else if (typeof Request !== "undefined" && input instanceof Request) raw = input.url;
+  if (!raw) return undefined;
+  try {
+    const parsed = new URL(raw, "http://cellfence.local");
+    if (parsed.protocol !== "http:" && parsed.protocol !== "https:") return undefined;
+  } catch {
+    return undefined;
+  }
+  return raw;
 }
 
 // 0.4.x (N-13): snapshot the disable flag at module-load time.
diff --git a/tests/trace.test.mjs b/tests/trace.test.mjs
index 1f21e84..9a16788 100644
--- a/tests/trace.test.mjs
+++ b/tests/trace.test.mjs
@@ -558,6 +558,43 @@ test("trace hook ignores fetch inputs that do not expose a URL selector", () =>
   assert.deepEqual(evidence.accesses, []);
 });
 
+test("trace hook ignores fetch calls with non-HTTP schemes like data: URLs", () => {
+  const tempDir = fs.mkdtempSync(path.join(os.tmpdir(), "cellfence-trace-fetch-non-http-"));
+  fs.writeFileSync(path.join(tempDir, "app.mjs"), `
+    try {
+      await fetch("data:text/plain,hello");
+    } catch {}
+    try {
+      await fetch(new URL("data:application/json,%7B%22ok%22%3Atrue%7D"));
+    } catch {}
+    try {
+      await fetch("https://example.invalid/valid-http");
+    } catch {}
+  `);
+
+  const evidencePath = path.join(tempDir, "resource-evidence.json");
+  const result = spawnSync(process.execPath, [
+    "--import",
+    pathToFileURL(tracePath).href,
+    "app.mjs",
+  ], {
+    cwd: tempDir,
+    encoding: "utf8",
+    env: {
+      ...process.env,
+      CELLFENCE_TRACE_CELL: "runtime",
+      CELLFENCE_TRACE_OUT: evidencePath,
+    },
+  });
+
+  assert.equal(result.status, 0, result.stderr);
+  const evidence = JSON.parse(fs.readFileSync(evidencePath, "utf8"));
+  assert.equal(evidence.transcriptStatus, "active");
+  assert.deepEqual(evidence.accesses.map((access) => access.selector), [
+    "https://example.invalid/valid-http",
+  ]);
+});
+
 test("trace hook covers default cell/output and fd based skips", () => {
   const tempDir = fs.mkdtempSync(path.join(os.tmpdir(), "cellfence-trace-defaults-"));
   fs.mkdirSync(path.join(tempDir, "data"), { recursive: true });