Agent Skills: ticket-verify

  <substep n="1a" title="Track Progress">
    <action if="first iteration">Create todo list from all instruction step titles (TodoWrite)</action>
    <action>Mark current step as in_progress (TodoWrite)</action>
  </substep>

  <substep n="1b" title="Handle Step Attributes">
    <check>If optional="true" and NOT YOLO -> Use Question tool to ask user whether to include (Yes/Skip)</check>
    <check>If if="condition" -> Evaluate condition</check>
    <check>If for-each="item" -> Repeat step for each item</check>
    <check>If repeat="n" -> Repeat step n times</check>
  </substep>

  <substep n="1c" title="Execute Step Content">
    <action>Process step instructions (markdown or XML tags)</action>
    <action>Replace {{variables}} with values (ask user if unknown)</action>
    <execute-tags>
      <tag>action -> Perform the action</tag>
      <tag>action type="script" -> Execute as shell command, capture output</tag>
      <tag>check if="condition" -> Conditional block wrapping actions (requires closing &lt;/check&gt;)</tag>
      <tag>ask -> Prompt user using Question tool and WAIT for response. When options exist, present them as selectable choices. For open-ended questions, provide helpful suggested options with custom input enabled</tag>
      <tag>invoke-skill name="skill-name" -> Execute another skill via OpenCode skill system</tag>
      <tag>invoke-agent name="subagent" prompt="task" -> Delegate to subagent via Task tool</tag>
      <tag>goto step="x" -> Jump to specified step (used for explicit loop-back in Q-R style flows)</tag>
    </execute-tags>
  </substep>

  <substep n="1d" title="Handle template-output Tags">
    <if tag="template-output">
      <mandate>Generate content for this section</mandate>
      <mandate>Save to file (Write first time, Edit subsequent)</mandate>
      <action>Display generated content</action>
      <action>Use Question tool to present checkpoint options. WAIT for response.
        <options>
          <option label="Continue" description="Approve this section and move to the next step" />
          <option label="Advanced Elicitation" description="Explore this section deeper before proceeding" />
          <option label="YOLO" description="Auto-complete the rest of this workflow without stopping" />
        </options>
        <if response="Advanced Elicitation">
          <action>Invoke advanced-elicitation skill</action>
        </if>
        <if response="Continue">
          <action>Continue to next step</action>
        </if>
        <if response="YOLO">
          <action>Enter YOLO mode for the rest of the workflow</action>
        </if>
      </action>
    </if>
  </substep>

  <substep n="1e" title="Step Completion">
    <action>Mark current step as completed (TodoWrite)</action>
    <check>If no special tags and NOT YOLO:</check>
    <action>Use Question tool to ask:
      <options>
        <option label="Continue" description="Proceed to the next step" />
        <option label="Edit" description="Revise this step before moving on" />
      </options>
    </action>
  </substep>
</step>

<step n="2" title="Completion">
  <check>Confirm all instruction steps have been executed</check>
  <output>Workflow complete.</output>
</step>

<output>
  A verified ticket at status: ready (or status: draft with open
  findings if verification did not pass).

  The questions file accumulates verify findings as a final
  section, capturing what was reviewed, what was found, what was
  resolved, and what was waived. This is the audit trail for
  future ticket-verify re-runs or implementation-QRDSPIV's V-phase.

  Bundle state after this skill: verified and in the backlog
  (status: ready), or in continuation pending fixes.
</output>

<key-responsibility>
  Per spec §10, V is a two-stage review:
  - **Self-review (always)**: this skill runs the severity ladder
    against the ticket regardless of tag.
  - **Reviewer subagent (conditional)**: for primary tags feature /
    exploration / infrastructure, dispatch the ticket-reviewer
    subagent (already authored — read-only, fresh-context).
    Skip the subagent for bug / hotfix / chore (lower stakes;
    self-review suffices per the spec).

  Finding resolution per spec §10.2:
  - Self-evident → apply via Edit directly
  - Ambiguous → Question tool per finding for user approval
  - Structural → loop back to earlier phase (no mini-QRDSPIV)

  The draft→ready gate is the final firm checkpoint. User confirms;
  status transitions; ticket enters the backlog. The user can
  always decline transition and leave the ticket in draft for
  further iteration.
</key-responsibility>

<step n="1" title="Load Bundle">
  <action>Load and execute: references/steps/step-01-load-bundle.xml</action>
  <note>
    Read the ticket, research, questions, linked DRs. Also load
    references/severity-ladder.md into working memory — the
    self-review and finding resolution depend on it.
  </note>
</step>

<step n="2" title="Self-Review">
  <action>Load and execute: references/steps/step-02-self-review.xml</action>
  <note>
    Apply the severity ladder against the ticket. Walk through
    each ticket section comparing to the ladder's example
    defects. Produce a self-review finding report.
  </note>
</step>

<step n="3" title="Dispatch Reviewer Subagent (Conditional)">
  <action>Load and execute: references/steps/step-03-dispatch-reviewer.xml</action>
  <note>
    For primary tags feature / exploration / infrastructure,
    dispatch the ticket-reviewer subagent via Task tool. Wait for
    its finding report. Merge with the self-review findings,
    deduplicating where both audiences flagged the same issue.
  </note>
</step>

<step n="4" title="Resolve Findings">
  <action>Load and execute: references/steps/step-04-resolve-findings.xml</action>
  <note>
    Walk through the merged finding list in priority order
    (Critical → High → Medium → Low). For each finding:
    self-evident fix → Edit directly; ambiguous fix → Question
    tool for user approval; structural finding → identify loop-back
    target and exit cleanly (user re-enters earlier phase, then
    re-runs ticket-verify).
  </note>
</step>

<step n="5" title="Transition and Handoff" critical="true">
  <action>Load and execute: references/steps/step-05-transition-and-handoff.xml</action>
  <note>
    Final checkpoint. Verdict: BLOCK / CONDITIONAL / PASS.
    On PASS or CONDITIONAL-with-waivers-applied, present the
    draft→ready gate. User confirms; frontmatter status updates;
    last_updated stamped; ticket enters backlog.
  </note>
</step>

  1. **Verified to ready**: ticket is now in the backlog, available
     for implementation-QRDSPIV pickup. Bundle status reflects
     this. Present the user with bundle paths and a closing note.

  2. **Findings unresolved, still draft**: the verify-phase
     surfaced blocking issues the user did not approve resolution
     for. Ticket remains at status: draft. User can re-invoke
     ticket-verify after addressing the open findings (which may
     require looping back to an earlier phase).

  In both cases, the questions file logs the verify session for
  future audit.
</action>