diff --git a/docs/developer-testing-guidelines.md b/docs/developer-testing-guidelines.md
index 291ff5d..d85c670 100644
--- a/docs/developer-testing-guidelines.md
+++ b/docs/developer-testing-guidelines.md
@@ -13,6 +13,25 @@ credentials.** Every test mocks the HTTP client and the keychain/auth lookup.
 
 ## 1. Standard test commands
 
+### Canonical runner: `./run-tests.sh`
+
+The canonical full-validation command is the root-level runner. It invokes the
+project virtualenv interpreter and passes any extra arguments straight through
+to `pytest`:
+
+```bash
+# Full validation
+./run-tests.sh
+
+# Focused validation (extra args forward to pytest)
+./run-tests.sh tests/test_mcp_server.py -q
+```
+
+`run-tests.sh` runs `venv/bin/python -m pytest "$@"` and fails with a clear
+setup message if the virtualenv Python is missing (so a session never silently
+falls back to the wrong interpreter). The explicit `venv/bin/python -m pytest`
+forms below remain valid and equivalent.
+
 The test suite needs the project virtualenv (it provides the MCP SDK):
 
 ```bash
diff --git a/run-tests.sh b/run-tests.sh
new file mode 100755
index 0000000..6930108
--- /dev/null
+++ b/run-tests.sh
@@ -0,0 +1,13 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+PYTHON="$ROOT_DIR/venv/bin/python"
+
+if [[ ! -x "$PYTHON" ]]; then
+  echo "ERROR: expected virtualenv Python at $PYTHON" >&2
+  echo "Create the venv first, then run: venv/bin/python -m pytest" >&2
+  exit 1
+fi
+
+exec "$PYTHON" -m pytest "$@"
diff --git a/tests/test_run_tests_script.py b/tests/test_run_tests_script.py
new file mode 100644
index 0000000..1ba5512
--- /dev/null
+++ b/tests/test_run_tests_script.py
@@ -0,0 +1,65 @@
+"""Contract checks for the root-level `run-tests.sh` convenience runner (#473).
+
+`run-tests.sh` is the canonical full-validation entry point. It must invoke the
+project virtualenv interpreter, forward extra args to pytest, fail closed when
+the venv Python is missing, and stay repo-local (no network, no lock files).
+These checks pin that behavior so it cannot silently regress, and confirm the
+developer testing guide names the runner.
+"""
+import stat
+from pathlib import Path
+
+REPO_ROOT = Path(__file__).resolve().parent.parent
+SCRIPT = REPO_ROOT / "run-tests.sh"
+GUIDE = REPO_ROOT / "docs" / "developer-testing-guidelines.md"
+
+
+def _script_text() -> str:
+    return SCRIPT.read_text(encoding="utf-8")
+
+
+def test_run_tests_script_exists():
+    assert SCRIPT.is_file(), "run-tests.sh must exist at the repository root"
+
+
+def test_run_tests_script_is_executable():
+    mode = SCRIPT.stat().st_mode
+    assert mode & stat.S_IXUSR, "run-tests.sh must be executable (chmod +x)"
+
+
+def test_run_tests_script_has_strict_bash_flags():
+    text = _script_text()
+    assert text.startswith("#!/usr/bin/env bash"), "must use the bash shebang"
+    assert "set -euo pipefail" in text, "must set -euo pipefail"
+
+
+def test_run_tests_script_uses_venv_pytest_and_forwards_args():
+    text = _script_text()
+    # Resolves the venv interpreter relative to the script's own directory.
+    assert "venv/bin/python" in text, "must use the project virtualenv Python"
+    assert "-m pytest" in text, "must run pytest via the module"
+    assert '"$@"' in text, "must forward extra CLI args to pytest"
+
+
+def test_run_tests_script_fails_closed_without_venv():
+    text = _script_text()
+    # Missing venv must be an explicit, non-zero-exit error, not a silent
+    # fallback to the wrong interpreter.
+    assert "if [[ ! -x" in text, "must guard on an executable venv Python"
+    assert "exit 1" in text, "must exit non-zero when the venv is missing"
+    assert "ERROR" in text, "must print a clear error message"
+
+
+def test_run_tests_script_stays_repo_local():
+    text = _script_text()
+    # No network calls, no lock-file writes from the runner itself.
+    for forbidden in ("curl", "wget", "gitea_issue_lock.json"):
+        assert forbidden not in text, f"runner must not reference {forbidden!r}"
+
+
+def test_guide_names_canonical_runner():
+    text = " ".join(GUIDE.read_text(encoding="utf-8").split())
+    assert "./run-tests.sh" in text, "testing guide must name ./run-tests.sh"
+    assert "./run-tests.sh tests/test_mcp_server.py -q" in text, (
+        "testing guide must show the focused-validation example"
+    )