leeovery Continue an in-progress epic. Shows full phase-by-phase state and routes to the appropriate phase skill.
⚠️ ZERO OUTPUT RULE: Do not narrate your processing. Produce no output until a step or reference file explicitly specifies display content. No "proceeding with...", no discovery summaries, no routing decisions, no transition text. Your first output must be content explicitly called for by the instructions.
Instructions
Load framework.md and follow its instructions as written.
Step 0: Initialisation
Output the next fenced block as markdown (not a code block):
# **`■ Continue Epic`**
→ Proceed to Step 1.
Step 1: Discovery State
!node .claude/skills/workflow-continue-epic/scripts/gateway.cjs
If the above shows a script invocation rather than discovery output, the dynamic content preprocessor did not run. Execute the script before continuing:
node .claude/skills/workflow-continue-epic/scripts/gateway.cjs
If discovery output is already displayed, it has been run on your behalf.
Parse the discovery output to understand:
From the === EPICS (N) === section:
- one line per active epic —
{name}: {active_phases}(phases with items;(no phases)when none) count— the header count of active epics
From the === COMPLETED (N) === / === CANCELLED (N) === sections:
- one line per closed epic —
{name} (last phase: {phase}) completed_count/cancelled_count— the header counts
The per-epic state surface (all_done, reconcile_pending, analysis_caches, needs_sequencing, build_order_needs_sequencing, the discovery map) is the scoped dump Step 4 runs after validation; display and routing come from the view snapshot at Step 9.
IMPORTANT: Use ONLY this script for discovery. Do NOT run additional bash commands (ls, head, cat, etc.) to gather state.
→ Proceed to Step 2.
Step 2: Check Count and Arguments
If count is 0
Output the next fenced block as a code block:
No epics in progress.
Run /workflow-start to begin a new one.
STOP. Do not proceed — terminal condition.
If work_unit argument $0 provided
Store the work_unit.
→ Proceed to Step 4.
If work_unit not provided
→ Proceed to Step 3.
Step 3: Select Epic
Output the next fenced block as markdown (not a code block):
**`□ Select Epic`**
Output the next fenced block as markdown (not a code block):
> Showing your active epics for selection.
Load select-epic.md and follow its instructions as written.
→ On return, proceed to Step 4.
Step 4: Validate Selection
Load validate-selection.md and follow its instructions as written.
→ On return, proceed to Step 5.
Step 5: Backfill
node .claude/skills/workflow-legacy-research-split/scripts/detect.cjs {work_unit}
Parse qualifying_sources from the JSON output.
Then read discovery_map from the most recent discovery output and filter for rows where summary=absent or description=absent. Store the filtered list as items_to_recover.
If qualifying_sources is empty and items_to_recover is empty
→ Proceed to Step 6.
Otherwise
Load backfill-checks.md with work_unit = {work_unit}, qualifying_sources = {qualifying_sources}, items_to_recover = {items_to_recover}.
backfill-checks is terminal when recovery work landed — it commits and stops, advising the user to /clear and re-run /workflow-start. It returns only when nothing was written (the batch declined); the skipped items re-offer on the next entry.
→ On return, proceed to Step 6.
Step 6: Topic Discovery
Read analysis_caches from the most recent discovery output. Load topic-discovery-dispatch.md with work_unit = {work_unit}, analysis_caches = {analysis_caches}.
On return, new_arrivals is populated for Step 9 to render the callout.
→ On return, proceed to Step 7.
Step 7: Sequence Map
Read needs_sequencing from the most recent discovery output.
If needs_sequencing is true
Output the next fenced block as markdown (not a code block):
**`□ Sequence Map`**
Output the next fenced block as markdown (not a code block):
> Assigning a suggested execution order to the map's topics.
Load sequence-discovery-map.md with work_unit = {work_unit}.
On return, re-run discovery so the display sees the new order:
node .claude/skills/workflow-continue-epic/scripts/gateway.cjs {work_unit}
Hold the refreshed output as the most recent discovery output.
→ On return, proceed to Step 8.
Otherwise
→ Proceed to Step 8.
Step 8: Sequence Build Order
Read build_order_needs_sequencing from the most recent discovery output.
If build_order_needs_sequencing is true
Output the next fenced block as markdown (not a code block):
**`□ Sequence Build Order`**
Output the next fenced block as markdown (not a code block):
> Assigning the build order across the specification topics.
Load sequence-build-order.md with work_unit = {work_unit}.
→ On return, proceed to Step 9.
Otherwise
→ Proceed to Step 9.
Step 9: Display State and Menu
Output the next fenced block as markdown (not a code block):
**`□ Epic State`**
Output the next fenced block as markdown (not a code block):
> Showing the full phase-by-phase breakdown and available actions.
Load epic-display-and-menu.md with new_arrivals = {new_arrivals}.
→ On return, proceed to Step 10.
Step 10: Route Selection
Invoke the route stored for the user's selection — the selected ACTIONS entry's route from epic-display-and-menu.md (e.g. /workflow-discussion-entry epic {work_unit} {topic}). Selections with route (internal) resolve inside that reference and never reach this step.
Skills receive positional arguments: $0 = work_type (epic), $1 = work_unit, $2 = topic (when provided). The continue_discovery route hands to the discovery skill, which detects the existing work unit and re-shapes the map (existing-epic mode) — workflow-continue-epic navigates; discovery owns the shaping.
This skill ends. The invoked skill will load into context and provide additional instructions. Terminal.