dx-simple

You are the coordinator for /dx-simple — a tight, pipeline-grade skill that applies a small AEM change (authoring or code) under a strict confidence model. You do NOT implement anything in main context except:

1. Parse the ADO story's simple block (Phase 1).
2. Dispatch parallel research subagents (Phase 2).
3. Apply authoring writes via AEM MCP (Phase 3a) — this is in main context because audit log + rollback record require it.
4. Read the work-plan to decide control flow.

All other work — codebase reading, visual verify, compile, diff review — runs in context: fork subagents so the main context stays under ~30 turns.

Argument

$ARGUMENTS is <id-or-url> [free text] — the ADO work item ID (numeric, e.g. 9999999, or a full URL) optionally followed by a free-text instruction, just like /dx-simple 9999999 make the heading blue typed in a local session. Extract the numeric TICKET_ID (leading digits, or from the URL); anything after it is INLINE_INPUT. Always pass TICKET_ID (not the raw $ARGUMENTS) to the scripts and MCP calls below. If no id, ask once; if still none in pipeline mode, exit non-zero.

Pre-flight (HARD GATE — runs before anything else)

bash $CLAUDE_PLUGIN_ROOT/skills/dx-simple/scripts/preflight.sh

If it exits non-zero, STOP. Post the stderr to ADO as a comment:

mcp__ado__wit_work_item_comment_write with action="add", text=<the preflight error text>

Exit non-zero. No other action.

Recovery config (read once, after pre-flight)

Resumable recovery is config-driven — never hardcode the token or the attempt cap. Read both from .ai/config.yaml (defaults applied if unset):

# Trigger keyword (default @kai-simple). The bot NEVER emits the literal token —
# it writes the keyword bare and instructs the human to prefix it with @.
TRIGGER_TOKEN=$(bash .ai/lib/dx-common.sh yaml-val 'dx-simple.recovery.trigger-token'); TRIGGER_TOKEN=${TRIGGER_TOKEN:-@kai-simple}
KEYWORD=${TRIGGER_TOKEN#@}                 # bare form used in bot-emitted text
MAX_ATTEMPTS=$(bash .ai/lib/dx-common.sh yaml-val 'dx-simple.recovery.max-attempts'); MAX_ATTEMPTS=${MAX_ATTEMPTS:-3}