aitask-pickweb-remotelisted

## Overview This skill is a sandboxed version of `aitask-pickrem` designed specifically for **Claude Code Web**, where the environment has no push access to `aitask-locks`, `aitask-data`, or `main` branches. It combines task selection and implementation into a single self-contained workflow with **zero interactive prompts** and **zero cross-branch operations**. All task metadata (plans, completion markers) is stored in `.aitask-data-updated/` on the current working branch. A separate local skill (`aitask-web-merge`) handles merging code to main and archiving task data after the Claude Web session completes. **Key differences from `aitask-pickrem`:** - NO lock acquisition (`aitask_pick_own.sh`) — read-only lock check only - NO status updates (`aitask_update.sh`) — task status stays as-is - NO archival (`aitask_archive.sh`) — completion marker written instead - NO `./ait git` — uses regular `git` for commits - Plan stored in `.aitask-data-updated/` instead of `aiplans/` - Completion marker JSON signals to `aitask-web-merge` that the branch is ready ## Arguments **Required:** Task ID (first positional argument) - Format 1: Parent task number (e.g., `42`) - Format 2: Child task ID (e.g., `42_2`) **IMPORTANT:** This skill will NOT work without a task ID argument. If invoked without one, display an error and abort. ## Workflow ### Step 0a: Initialize Data Branch (if needed) Ensure the aitask-data worktree and symlinks are set up: ```bash ./.aitask-scripts/aitask_init_data