benbrastmckie/nvim/skill-epi-implement

Epi Implement Skill

Thin wrapper that delegates epidemiology implementation to epi-implement-agent subagent.

IMPORTANT: This skill implements the skill-internal postflight pattern. After the subagent returns, this skill handles all postflight operations (status update, artifact linking, git commit) before returning.

Context References

Reference (do not load eagerly):

• Path: .claude/context/formats/return-metadata-file.md - Metadata file schema
• Path: .claude/context/patterns/postflight-control.md - Marker file protocol
• Path: .claude/context/patterns/file-metadata-exchange.md - File I/O helpers
• Path: .claude/context/patterns/jq-escaping-workarounds.md - jq escaping patterns (Issue #1132)

Note: This skill is a thin wrapper with internal postflight. Context is loaded by the delegated agent.

Trigger Conditions

This skill activates when:

• /implement on a task with task_type: "epi" or task_type: "epi:study"
• Epidemiology extension is available

Input Parameters

Required Parameters

• task_number - Task number (must exist in state.json with task_type="epi" or "epi:study")
• session_id - Session ID from orchestrator

Execution Flow

Stage 1: Input Validation

Validate required inputs:

• task_number - Must be provided and exist in state.json
• Verify task_type is "epi" or "epi:study"

task_data=$(jq -r --argjson num "$task_number" \
  '.active_projects[] | select(.project_number == $num)' \
  specs/state.json)
 
if [ -z "$task_data" ]; then
  return error "Task $task_number not found"
fi
 
task_type=$(echo "$task_data" | jq -r '.task_type // ""')
status=$(echo "$task_data" | jq -r '.status')
project_name=$(echo "$task_data" | jq -r '.project_name')
description=$(echo "$task_data" | jq -r '.description // ""')
 
if [ "$task_type" = "epi" ] || [ "$task_type" = "epi:study" ]; then
  : # valid
else
  return error "Task $task_number is not an epi task (task_type=$task_type)"
fi

Stage 2 + Stage 3: Preflight Status Update and Postflight Marker

Source skill-base.sh once, then follow @.claude/context/patterns/skill-preflight-flow.md in full for Stage 2 (preflight status update) and Stage 3 (marker creation):

source .claude/scripts/skill-base.sh
padded_num=$(printf "%03d" "$task_number")
skill_name="skill-epi-implement"
operation="implement"

operation="implement" (not "epi_implement") is required here: update-task-status.sh's target_status vocabulary has no epi_implement value, so this skill maps onto the plain implement operation, same as skill-orchestrate. The marker's operation field now reads "implement" rather than "epi_implement" — a byte-for-byte Shape A key set, matching what this skill's marker already carried (created, stop_hook_active) before this conversion.

Stage 4a: Memory Retrieval and Literature Detection

Skip memory retrieval if: clean_flag is true (from --clean).

if [ "$clean_flag" != "true" ]; then
  memory_context=$(bash .claude/scripts/memory-retrieve.sh "$description" "$task_type" "" 2>/dev/null) || memory_context=""
fi

Follow @.claude/context/patterns/lit-stage4a-flow.md in full to resolve --lit and set lit_context, exactly as skill-orchestrate does. This skill supplies the shared block's preconditions: lit_flag, description, orchestrator_mode (default "false" when unset).

Stage 4: Prepare Delegation Context

Read the implementation plan to identify phases for per-phase delegation.

{
  "session_id": "sess_{timestamp}_{random}",
  "delegation_depth": 1,
  "delegation_path": ["orchestrator", "implement", "skill-epi-implement"],
  "timeout": 3600,
  "task_context": {
    "task_number": N,
    "task_name": "{project_name}",
    "description": "{description}",
    "task_type": "{task_type}"
  },
  "workflow_type": "epi_implement",
  "forcing_data": "{from state.json task metadata: study_design, data_paths, etc.}",
  "research_report_path": "specs/{NNN}_{SLUG}/reports/{MM}_epi-research.md",
  "plan_path": "specs/{NNN}_{SLUG}/plans/{MM}_implementation-plan.md",
  "metadata_file_path": "specs/{NNN}_{SLUG}/.return-meta.json"
}

Phased delegation: Read plan phases from the plan file. Delegate each phase to the subagent sequentially, passing phase number and details. Resume from last incomplete phase on re-invocation.

Stage 5: Invoke Subagent

CRITICAL: Use the Agent tool to spawn the subagent.

Tool: Agent (NOT Skill, NOT Plan)
Parameters:
  - subagent_type: "epi-implement-agent"
  - prompt: [Include task_context, delegation_context, workflow_type, forcing_data, research_report_path, plan_path, metadata_file_path]
  - description: "Execute epi_implement for task {N}"

DO NOT use Skill(epi-implement-agent) - this will FAIL.

For phased plans, invoke subagent per-phase with phase context. Commit after each phase.

Stage 5b: Self-Execution Fallback

CRITICAL: If you performed the work above WITHOUT using the Agent tool (i.e., you read files, wrote artifacts, or updated metadata directly instead of spawning a subagent), you MUST write a .return-meta.json file now before proceeding to postflight. Use the schema from return-metadata-file.md with the appropriate status value for this operation.

If you DID use the Agent tool, skip this stage -- the subagent already wrote the metadata.

Postflight (ALWAYS EXECUTE)

The following stages MUST execute after work is complete, whether the work was done by a subagent or inline (Stage 5b). Do NOT skip these stages for any reason.

Stage 6: Read Metadata File

metadata_file="specs/${padded_num}_${project_name}/.return-meta.json"
 
if [ -f "$metadata_file" ] && jq empty "$metadata_file" 2>/dev/null; then
    status=$(jq -r '.status' "$metadata_file")
    artifact_path=$(jq -r '.artifacts[0].path // ""' "$metadata_file")
    artifact_type=$(jq -r '.artifacts[0].type // "summary"' "$metadata_file")
    artifact_summary=$(jq -r '.artifacts[0].summary // ""' "$metadata_file")
    memory_candidates=$(jq -c '.memory_candidates // []' "$metadata_file")
    # Schema: .claude/context/formats/return-metadata-file.md
    completion_summary=$(jq -r '.completion_data.completion_summary // ""' "$metadata_file")
    roadmap_items=$(jq -c '.completion_data.roadmap_items // []' "$metadata_file")
else
    echo "Error: Invalid or missing metadata file"
    status="failed"
fi

Note: renamed the local variable from meta_status to status to match the skill-postflight-flow.md shared block's precondition naming used below.

Stage 7, 7a, 8, 8a: Postflight Status, Memory Candidates, Artifact Linking, Notify

MUST NOT cross the postflight boundary: The subagent MUST NOT update state.json or TODO.md status. Only this skill performs postflight status transitions.

Follow @.claude/context/patterns/skill-postflight-flow.md for Stage 7 (postflight status update), Stage 7a (memory-candidate propagation), Stage 8 (artifact linking; artifact type "summary", an implementation summary with R script paths), and Stage 8a (TTS notify). skill_propagate_completion_summary (already routed through skill-base.sh before this conversion, sourced once at Stage 2 + Stage 3 above) is preserved as a "not covered by the shared block" call:

field_name='**Summary**'
next_field='**Description**'
skill_postflight_update "$task_number" "$operation" "$session_id" "$status"
if [ "$status" = "implemented" ] || [ "$status" = "completed" ]; then
    skill_propagate_completion_summary "$task_number" "$completion_summary" "$roadmap_items" "$task_type"
fi
skill_propagate_memory_candidates "$task_number" "$memory_candidates" "$session_id"
skill_link_artifacts "$task_number" "$artifact_path" "$artifact_type" "$artifact_summary" \
  "$field_name" "$next_field" "$session_id"
skill_lifecycle_notify "$status"

Stage 9: Git Commit

Apply the implement scope from .claude/context/standards/git-staging-scope.md — targeted staging, never a repo-wide add:

stage_paths=("specs/${padded_num}_${project_name}/" "specs/TODO.md" "specs/state.json")
metadata_file="specs/${padded_num}_${project_name}/.return-meta.json"
while IFS= read -r f; do
  [ -n "$f" ] && stage_paths+=("$f")
done < <(jq -r '.modified_files[]? // empty' "$metadata_file" 2>/dev/null)
bash .claude/scripts/git-commit-scoped.sh \
  --message "task ${task_number}: complete epi implementation" \
  --session "${session_id}" \
  --honest-index-rows "${task_number}" \
  -- "${stage_paths[@]}"

Stage 10: Cleanup

Follow @.claude/context/patterns/skill-postflight-flow.md's Stage 9 (cleanup):

skill_cleanup "$padded_num" "$project_name"

Stage 11: Return Brief Summary

Implementation Success:

Epi implementation completed for task {N}:
- R analysis scripts created and validated
- Created summary at specs/{NNN}_{SLUG}/summaries/{MM}_epi-implementation-summary.md
- Status updated to [COMPLETED]
- Changes committed with session {session_id}

Error Handling

Task not found

Epi implement skill error for task {N}:
- Task not found in state.json
- No status changes made

Wrong task_type

Epi implement skill error for task {N}:
- Task is not an epi task (task_type={task_type})
- No status changes made

Metadata file missing

Keep status at preflight level for resume.

Git commit failure

Non-blocking. Log failure but continue.

MUST NOT (Postflight Boundary)

After the agent returns -- whether with status implemented, partial, or failed -- this skill MUST proceed immediately to postflight (Stage 6). The skill MUST NOT:

1. Edit R/Python files - All analysis work is done by agent
2. Run R scripts - Execution is done by agent
3. Use MCP tools - Domain tools are for agent use only
4. Analyze or grep source - Analysis is agent work
5. Write summary/reports - Artifact creation is done by agent

PROHIBITION: If the subagent returned partial or failed status, the lead skill MUST NOT attempt to continue, complete, or "fill in" the subagent's work. Report the partial/failed status and let the user re-run /implement to resume.

The postflight phase is LIMITED TO:

• Reading agent metadata file
• Updating state.json via jq
• Updating TODO.md status marker via Edit
• Linking artifacts in state.json
• Git commit
• Cleanup of temp/marker files

Reference: @.claude/context/standards/postflight-tool-restrictions.md

Return Format

This skill returns a brief text summary (NOT JSON). The JSON metadata is written to the file and processed internally.