Add tests for todo completion, task duplication, and execution limits

- Implement test for finalizing a linked ready todo and ensuring it stays archived after task reconciliation refresh.
- Add test to verify that a local template task duplicates correctly when the stored fallback prompt is empty.
- Introduce test to check that deleting a task restores task and job state when persist fails.
- Add tests to ensure that executing due tasks and running tasks now skips tasks that are already in flight.
- Update todo completion tests to auto-dismiss inline confirm after 30 seconds and clear timers on cancel and confirm.
- Modify todo editor completion to use inline confirm with a 30-second reset.
- Enhance request handling for approve and finalize todo actions to confirm before dispatching.
- Ensure SQLite database writes through a temp-file rename instead of overwriting in place.
- Update types to include new message types for requestApproveTodo and requestFinalizeTodo.

Author: goodguy1963

.github/agents/ceo.agent.md

@@ -3,7 +3,7 @@ description: Strategic orchestrator that merges into repo-local agent systems, d
 name: CEO
 argument-hint: Ask me to coordinate work, review a direction, route to specialists, or evolve the repo's agent system.
 model: GPT-5.4 (copilot)
-tools: [vscode/memory, execute/runNotebookCell, execute/testFailure, execute/executionSubagent, execute/getTerminalOutput, execute/killTerminal, execute/sendToTerminal, execute/runTask, execute/createAndRunTask, read/readFile, agent/runSubagent, search/codebase, search/listDirectory, search/textSearch, scheduler/cockpit_get_board]
+tools: [vscode/memory, execute/runNotebookCell, execute/executionSubagent, execute/getTerminalOutput, execute/killTerminal, execute/sendToTerminal, execute/runTask, execute/createAndRunTask, read/readFile, agent/runSubagent, search/codebase, search/listDirectory, search/textSearch, perplexity/perplexity_ask, perplexity/perplexity_reason, perplexity/perplexity_research, perplexity/perplexity_search, scheduler/cockpit_get_board]
 handoffs:
   - label: Plan Work
     agent: Planner

.vscodeignore

@@ -29,6 +29,21 @@ AGENTS.md
 !.github/prompts/**
 !.github/skills/
 !.github/skills/**
+.artifacts/**
+compile.log
+compile*.log
+patch.diff
+plan.md
+test.js
+test.txt
+test_output.log
+test_output.txt
+test_output*.log
+test_output*.txt
+ts_errors.log
+ts_errors.txt
+ts_errors*.log
+ts_errors*.txt
 
 # Dev-only artifacts
 output_sessions/**

README.md

@@ -16,15 +16,10 @@
     </a>
 </p>
 
-<p align="center">
-    Workflow status for the release pipeline · Stable = latest tagged production release · Edge = rolling prerelease built from <code>main</code>
-</p>
 <!-- markdownlint-enable MD033 MD041 -->
 
 Copilot Cockpit helps you plan AI work, approve it, and then run it with visible checkpoints instead of handing your repo to a blind autonomous loop.
 
-The strongest demo is not claiming the repo can run itself. It is showing bounded recurring work that a person would actually keep: scouting opportunities, checking delivery risk, packaging knowledge, and then stopping for review.
-
 ![AI orchestration core loop](images/AI%20Orchestration%20Core%20Loop.png)
 
 

docs/feature-tour.md

@@ -135,7 +135,6 @@ Caption: Start with the operating model.
 
 - Start there if you are opening the extension for the first time.
 - Use it to understand the planning versus execution model before scheduling anything.
-- You can reopen the same guided flow from the top bar with `Intro Tutorial`, and use the top-bar `Plan Integration` button when you are ready to inspect repo-local agent surfaces.
 
 Best for: first-time users who want the operating model before the controls.