Reverse engineering tools per mode; raw firmware in decompile; skill

- AgentConfig.reverse_tools (Settings -> Agents & commands) gives any mode
  binary_info and decompile; the clean-room check of the tools applies to
  craft only (ToolCtx.clean_room = clean room && craft).
- decompile: optional processor (Ghidra language id) and base_address for raw
  images (BinaryLoader); separate cache per processor/address.
- Skill "reverse-engineering": identify, convert HEX, processor table for
  common chips, decompile, explore, notes in analysis/.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

web/src/i18n/en.ts

@@ -188,7 +188,7 @@ export const en = {
   "tools.install": "Install Ghidra",
   "tools.reinstall": "Check / repair installation",
   "tools.installing": "Installing (download ~800 MB, on ARM the decompiler is compiled – this takes several minutes)…",
-  "tools.ghidraHint": "Ghidra lets agents in craft Deep mode decompile programs (Windows .exe/.dll, Linux, DOS, …) to study how they work. Installed into the data volume together with its own JDK.",
+  "tools.ghidraHint": "Ghidra lets agents decompile programs and firmware (Windows .exe/.dll, Linux, DOS, …) to study how they work – in craft Deep mode and in modes with reverse engineering tools (Settings → Agents & commands). Installed into the data volume together with its own JDK.",
   "rewind.intro": "The conversation is reset to before this message; you can edit and resend it.",
   "rewind.noFiles": "The agent did not change files since then.",
   "rewind.restore": "Also restore the {n} file(s) the agent changed",
@@ -335,6 +335,8 @@ export const en = {
   "activity.reviewing": "Reviewing the result (round {round})…",
   "chat.reviewPassed": "Review passed:",
   "chat.todoReminder": "Back to work – {n} open todos:",
+  "agents.reverseTools": "Reverse engineering tools (binary_info, decompile with Ghidra)",
+  "agents.reverseToolsHint": "The agent can inspect and decompile programs and firmware in this mode (Ghidra must be installed under Settings → Tools). Craft has them anyway in deep mode.",
   "agents.refineAfterTodos": "Review every completed todo and work until the list is done",
   "agents.refineAfterTodosHint": "Each todo the agent marks completed is reviewed; with findings it is reopened. Ending the turn with open todos sends the agent back to work (for unattended runs, e.g. overnight). Off: only one review at the end of the turn.",
   "chat.workSteps": "{n} steps",