Move the code and terraform audits into the reviews plugin

Copy the standalone code-review and terraform-review skills into
plugins/reviews as audit-code and audit-terraform. The rename separates the
automated, linter-driven audits from the guided review-pr walkthrough that
already lived here.

Resolve bundled script paths through ${SKILL_DIR}, exported in a new step 0.
CLAUDE_PLUGIN_ROOT is not set in the Bash tool environment, so the obvious
substitution would have expanded to nothing and broken every collection
script invocation.

Replace the PLAN and DESIGN docs with READMEs written from the current
SKILL.md and scripts. The old docs had drifted badly: they named semgrep
where the code calls opengrep, scoped five review agents where there are
now eight, and predated Lua, PowerShell, and GitHub Actions support.

Add CONSISTENCY_NORMS to the audit-terraform agent inputs. The collection
script writes consistency_norms.json and the agent prompt declares it, but
SKILL.md never listed it, leaving the variable unsubstituted.

Drop the --ingest-verdicts instruction from both skills. review_stats.py
parses no arguments, so the ref-mode verdict template it told users to feed
back could never be read.

Point audit-terraform's smoke test at README.md and resolve its fixture
paths relative to the test file rather than an absolute home directory.

Tests: 197 passing (audit-code), 106 passing (audit-terraform).

plugins/reviews/skills/audit-code/tests/test_runner.py

@@ -0,0 +1,107 @@
+from unittest.mock import patch, MagicMock
+
+from scripts.runner import (
+    run_tool, tool_available, run_tools_parallel,
+)
+
+
+def test_tool_available_returns_true_when_on_path():
+    with patch("shutil.which", return_value="/usr/bin/bandit"):
+        assert tool_available("bandit") is True
+
+
+def test_tool_available_returns_false_when_missing():
+    with patch("shutil.which", return_value=None):
+        assert tool_available("nope") is False
+
+
+def test_run_tool_captures_stdout():
+    fake = MagicMock(returncode=0, stdout='{"results": []}', stderr="")
+    with patch("subprocess.run", return_value=fake):
+        with patch("shutil.which", return_value="/usr/bin/bandit"):
+            r = run_tool(["bandit", "-r", "."], cwd="/repo")
+    assert r.ran is True
+    assert r.exit_code == 0
+    assert r.stdout == '{"results": []}'
+
+
+def test_run_tool_reports_missing_binary():
+    with patch("shutil.which", return_value=None):
+        r = run_tool(["nope"], cwd="/repo")
+    assert r.ran is False
+    assert r.reason == "not on PATH"
+
+
+def test_run_tool_reports_timeout():
+    import subprocess
+    def _raise(*a, **kw):
+        raise subprocess.TimeoutExpired(cmd="x", timeout=1)
+    with patch("subprocess.run", side_effect=_raise):
+        with patch("shutil.which", return_value="/usr/bin/x"):
+            r = run_tool(["x"], cwd="/repo", timeout=1)
+    assert r.ran is False
+    assert "timeout" in r.reason.lower()
+
+
+def test_run_tools_parallel_respects_cap():
+    def _fake(args, **kw):
+        return MagicMock(returncode=0, stdout="", stderr="")
+    with patch("subprocess.run", side_effect=_fake):
+        with patch("shutil.which", return_value="/usr/bin/x"):
+            jobs = [(["a"], "/r"), (["b"], "/r"), (["c"], "/r")]
+            results = run_tools_parallel(jobs, max_workers=2)
+    assert len(results) == 3
+    assert {r.binary for r in results} == {"a", "b", "c"}
+
+
+def test_resolve_tool_prefers_skill_venv(tmp_path):
+    """If a tool exists in the skill's .venv/bin/, prefer it over PATH."""
+    from scripts import runner
+
+    venv_bin = tmp_path / "venv" / "bin"
+    venv_bin.mkdir(parents=True)
+    bandit_local = venv_bin / "bandit"
+    bandit_local.write_text("#!/bin/sh\necho hi\n")
+    bandit_local.chmod(0o755)
+
+    with patch.object(runner, "_SKILL_VENV_BIN", venv_bin):
+        with patch("shutil.which", return_value="/usr/bin/bandit"):
+            resolved = runner.resolve_tool("bandit")
+    assert resolved == str(bandit_local)
+
+
+def test_resolve_tool_falls_back_to_path(tmp_path):
+    """Without a local venv copy, fall back to PATH."""
+    from scripts import runner
+
+    venv_bin = tmp_path / "venv" / "bin"
+    venv_bin.mkdir(parents=True)
+
+    with patch.object(runner, "_SKILL_VENV_BIN", venv_bin):
+        with patch("shutil.which", return_value="/usr/bin/bandit"):
+            resolved = runner.resolve_tool("bandit")
+    assert resolved == "/usr/bin/bandit"
+
+
+def test_run_tool_uses_resolved_path(tmp_path):
+    """run_tool should invoke the resolved binary, not the bare name."""
+    from scripts import runner
+
+    venv_bin = tmp_path / "venv" / "bin"
+    venv_bin.mkdir(parents=True)
+    bandit_local = venv_bin / "bandit"
+    bandit_local.write_text("#!/bin/sh\necho hi\n")
+    bandit_local.chmod(0o755)
+
+    captured: list[list[str]] = []
+
+    def _fake(args, **kw):
+        captured.append(list(args))
+        return MagicMock(returncode=0, stdout="", stderr="")
+
+    with patch.object(runner, "_SKILL_VENV_BIN", venv_bin):
+        with patch("subprocess.run", side_effect=_fake):
+            runner.run_tool(["bandit", "-r", "."], cwd="/repo")
+
+    assert captured[0][0] == str(bandit_local)
+    assert captured[0][1:] == ["-r", "."]