rpa-bot-handoff: one short Handover SOP (max 4 pages) instead of the 11-section guide

The first real handover (MEK Charge Reconciliation) came out at 9 pages.
Ops want it short: which URL to open and what to click, what shows where
after the click, what to do when an exception comes, and how to email us.

- Output is "<Project Name> Handover SOP" (HANDOVER-SOP.md + .docx) with
  5 sections: how to run, what you see while it runs, result, when
  something goes wrong, email us. HANDOVER-DEV.md is dropped; dev-team
  findings go in the chat.
- At most ~1,400 words; to_docx.py prints the word count and flags overflow,
  and renders *italic* for the variable parts of messages.
- Every quoted message, label and URL is checked against the code. The SOP
  must say what clicking Run again does to the data.
- One scan subagent; failure classes stay as a scan checklist.
- Tebra reference rewritten in the SOP format; test, README, glossary
  updated; ADR-0005 supersedes ADR-0004.

The handoff eval runner (tests/evals/run_handoff_eval.py) still grades an
older 13-section format and was already out of date; not touched here.

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

skills/rpa-bot-handoff/templates/HANDOVER-SOP.md

@@ -0,0 +1,73 @@
+# {{PROJECT_NAME}} Handover SOP
+
+Version {{VERSION}} · {{DATE}} · Problems: email {{EMAIL_GROUP}} (section 5)
+
+{{ONE_LINE: what the bot does, from which app or portal, and what it produces}}
+
+## 1. How to run
+
+1. {{e.g. Open https://... in Chrome / Open "<app>" from the Desktop.}}
+2. {{e.g. Sign in with the team password.}}
+3. {{e.g. Click "Run".}}
+4. {{e.g. Wait until the status says "Finished" or "Stopped" (section 2).}}
+
+**Before you click {{RUN_LABEL}}:**
+- {{WHAT A NEW CLICK DOES: e.g. "Every click starts a fresh run on today's data. It deletes the previous run's reports first, so download them before you click."}}
+- {{e.g. "Only one run at a time. A greyed-out button means a run is going."}}
+
+## 2. What you see while it runs
+
+| Where on the screen | What it shows |
+|---|---|
+| {{AREA, exact label}} | {{TEXTS IT SHOWS, in order}} |
+
+**A normal run:**
+1. {{e.g. Both lanes say "Signing in…".}}
+2. {{e.g. Each lane fills: "3 of 10 locations".}}
+3. {{e.g. Status "Finished", last line "Finished. Your reports are ready to download below."}}
+
+| Step | The bot waits at most | Then |
+|---|---|---|
+| {{STEP}} | {{e.g. 3 min}} | {{e.g. skips that location / stops / retries 3 times}} |
+
+**Step in only when:** {{e.g. the run has ended, or nothing new has appeared for 10 minutes}}.
+
+## 3. Result
+
+- **Good run:** {{HOW IT ENDS ON SCREEN}}
+- **Needs a look:** {{HOW IT ENDS, including any misleading green or "Finished"}}
+
+**Where:** {{e.g. Reports box → Download. Excel: Summary sheet, Reconciliation sheet, Result column.}}
+
+| Status | Meaning | What you do |
+|---|---|---|
+| {{STATUS}} | {{ONE LINE}} | {{ACTION or "nothing"}} |
+
+## 4. When something goes wrong
+
+**How it shows:** {{e.g. a red or orange dot in the activity list under the Run button; the lane turns red}}.
+
+| ID | Message you see | What it means | What you do |
+|---|---|---|---|
+| E01 | {{EXACT MESSAGE}} | {{ONE LINE}} | {{STEPS. "Email us" when ops can't fix it}} |
+
+**Don't:**
+- {{DON'T + short reason}}
+
+**To stop it:** {{THE ONE RIGHT WAY}}
+
+## 5. Email us
+
+Email **{{EMAIL_GROUP}}** when section 4 says "email us", or when the same error comes back after one re-run. Answer: {{RESPONSE_TIME}}.
+
+```
+Subject: {{PROJECT_NAME}} - E<number> - <date>
+
+Run started: <time from the screen>
+Error: E<number> - <the message, copied>
+Where: <location / item>
+What I did: <e.g. ran it again once>
+Attached: screenshot of the screen
+```
+
+**Never send:** passwords, {{SENSITIVE: e.g. the Excel reports (they hold patient names)}}.