/skill:jat-complete - Finish Task Properly

Complete current task with full verification protocol. Session ends after completion.

Usage

/skill:jat-complete             # Complete task, show completion block
/skill:jat-complete --kill (-k) # Complete and auto-kill session

If --kill or -k is in the invocation args, set IS_KILL=true for the rest of this flow (referenced in STEP 3.8 below).

What This Does

1. Verify task (tests, lint, security)
2. Commit changes with proper message
3. Write memory entry - Save context for future agents
4. Reply to reporter (if applicable) - Post a friendly customer-facing comment when the task has a requester/approver
5. Mark task complete (jt close)
6. Emit completion signal to IDE

Prerequisites

You MUST have emitted a review signal before running this:

jat-signal review '{
  "taskId": "TASK_ID",
  "taskTitle": "TASK_TITLE",
  "summary": ["What you accomplished"],
  "filesModified": [
    {"path": "src/file.ts", "changeType": "modified", "linesAdded": 50, "linesRemoved": 10}
  ]
}'

Step-by-Step Instructions

STEP 1: Get Current Task and Agent Identity

1A: Get Agent Name

Check the tmux session name or identity file:

TMUX_SESSION=$(tmux display-message -p '#S' 2>/dev/null)
# Agent name is the tmux session without "jat-" prefix
AGENT_NAME="${TMUX_SESSION#jat-}"

1B: Get Current Task

Find your in-progress task:

jt list --json | jq -r '.[] | select(.assignee == "AGENT_NAME" and .status == "in_progress") | .id'

If no task found, check for spontaneous work (uncommitted changes without a formal task).

STEP 1D: Spontaneous Work Detection

Only if no in_progress task was found.

Check git status and conversation context for work that was done without a formal task:

git status --porcelain
git diff --stat
git log --oneline -5

If work is detected, propose creating a backfill task record:

jt create "INFERRED_TITLE" \
  --type INFERRED_TYPE \
  --description "INFERRED_DESCRIPTION" \
  --assignee "$AGENT_NAME" \
  --status in_progress

If no work detected, exit the completion flow.

STEP 2: Verify Task

Run verification checks appropriate to the project:

# Emit verifying signal
jat-step verifying --task "$TASK_ID" --title "$TASK_TITLE" --agent "$AGENT_NAME"

# Then run checks:
# - Tests (npm test, pytest, etc.)
# - Lint (eslint, ruff, etc.)
# - Type check (tsc --noEmit, etc.)
# - Build (npm run build, etc.)

If verification fails, stop and fix issues before continuing.

STEP 2.1: Runtime Surface Check (MANDATORY)

svelte-check / type checking is type-only. Run the runtime gate too:

jat-runtime-verify --task "$TASK_ID"

This detects changed surfaces from the git diff and exercises the real runtime:

• API routes (ide/src/routes/api/**): rebuilds IDE, spins throwaway server on port 3399, hits each changed endpoint — fail on 5xx or connection refused
• CLI/tools (tools/**, cli/**): runs each changed command with --help — fail if unreachable
• UI (ide/src/lib/components/**, +page*, +layout*): advisory only — run /skill:jat-verify afterward
• Docs/config only: skips automatically

Exit 1 = gate fails. Do not continue to STEP 3. Fix the runtime issue and re-run.

The gate can be skipped ONLY for documentation-only changes (CLAUDE.md, AGENTS.md, skills/, commands/, shared/ — no code touched).

STEP 2.5: Update Documentation (If Appropriate)

Only update docs when changes affect how others use the codebase:

• New tool/command added
• New API endpoint
• Breaking change
• New configuration option

Most tasks do NOT need doc updates.

STEP 3: Commit Changes

# Get task type for commit prefix
TASK_TYPE=$(jt show "$TASK_ID" --json | jq -r '.[0].issue_type // "task"')

# Commit with proper message format
jat-step committing --task "$TASK_ID" --title "$TASK_TITLE" --agent "$AGENT_NAME" --type "$TASK_TYPE"

If jat-step is not available, commit manually:

# Stage ONLY the specific files you modified — never git add -A, git add ., or git add -u
git add path/to/file1 path/to/file2
git commit -m "TASK_TYPE($TASK_ID): TASK_TITLE

Co-Authored-By: Pi Agent <noreply@pi.dev>"

STEP 3.5: Write Memory Entry

Save context from this session for future agents. Use the Write tool to create:

.jat/memory/{YYYY-MM-DD}-{taskId}-{slug}.md

Include YAML frontmatter (task, agent, project, completed, files, tags, labels, priority, type) and sections: Summary, Approach, Decisions (if notable), Key Files, Lessons (if any).

Then trigger incremental index:

jat-memory index --project "$(pwd)"

If indexing fails, log the error but continue. Memory is non-blocking.

STEP 3.8: Reply to Reporter (Customer-Facing Comment)

Posts a customer-facing reply comment BEFORE close for tasks that originated from user feedback. Comments are the canonical channel per epic jat-47wul — they render in the IDE thread AND in the feedback widget back to the reporter (jat-47wul.4). The server forces external: true for author_type: agent per jat-47wul.2, so agent comments are always customer-visible.

KILL_FLAG=""
if [[ "$IS_KILL" == true ]]; then KILL_FLAG="--kill"; fi
jat-step replying --task "$TASK_ID" --title "$TASK_TITLE" --agent "$AGENT_NAME" $KILL_FLAG

jat-step replying generates a completion bundle, extracts the LLM-authored devResponse field, resolves the reporter from task identity (approver → requester → creator), and POSTs the reply via /api/tasks/:id/comments. The bundle is cached at /tmp/jat-bundle-<task>.json so STEP 6 reuses it without paying for a second LLM call.

The step is non-blocking. It silently exits 0 when:

• ANTHROPIC_API_KEY is unset or bundle generation fails
• The bundle has no devResponse (LLM determined task isn't user-originated)
• Task has no reporter email (approver/requester/creator all null)
• IDE is unreachable at localhost:3333

For internal chores, refactors, and dev-spawned tasks with no external reporter, no comment is posted and the close proceeds normally.

STEP 4: Mark Task Complete

jat-step closing --task "$TASK_ID" --title "$TASK_TITLE" --agent "$AGENT_NAME"

Or manually:

jt close "$TASK_ID" --reason "Completed by $AGENT_NAME"

STEP 4.5: Auto-Close Eligible Epics

Note: Integration callbacks (Supabase status sync, dev_notes) fire automatically from jt close — no agent action needed. Notes are sourced from: review signal → git commits → assignee name.

jt epic close-eligible

STEP 5: Emit Completion Signal

jat-step complete --task "$TASK_ID" --title "$TASK_TITLE" --agent "$AGENT_NAME"

This generates a structured completion bundle and emits the final complete signal.

Then output the completion banner:

TASK COMPLETED: $TASK_ID
Agent: $AGENT_NAME

Summary:
  - [accomplishment 1]
  - [accomplishment 2]

Quality: tests passing, build clean

Session complete. Spawn a new agent for the next task.

"Ready for Review" vs "Complete"

Never say "Task Complete" until jt close has run.

Error Handling

No task in progress:

No task in progress. Run /skill:jat-start to pick a task.

Verification failed:

Verification failed:
  - 2 tests failing
  - 5 lint errors
Fix issues and try again.

Step Summary