/lifetime

Every long-running process needs graceful shutdown. This skill maps shutdown readiness.

What this is for

• HTTP servers without graceful shutdown (connections dropped on SIGTERM)
• Workers/queues that don't drain jobs before exit
• Cron jobs with no rollback on interrupt
• Read-only skill. No configuration changes.

PROTECTION RULE - never ~/.claude/

Read-only skill. Guard required if write mode added later.

What You Must Do When Invoked

During analysis, assign a confidence level to each finding: proven (confirmed by evidence), likely (strong signal, needs review), or suspected (weak signal).

If /lifetime -help or /lifetime -h (without further arguments) is invoked: output the ## Usage section unchanged and stop.

Otherwise follow these steps in order, skipping none.

Step 1 - Help check

If invoked with -help or -h, output the ## Usage section unchanged and stop.

Step 2 - Clarify target

Clarify -ProjectDir. Get confirmation.

Step 3 - Scan

& "<SKILL_DIR>/scripts/lifetime-scan.ps1" -ProjectDir "<path>"

Step 4 - Classification

Read each process:

• Graceful: clean shutdown with drain, grace period, resource cleanup
• Rough: catches signal but no drain or grace period
• Abrupt: no signal handler, kills immediately
• Dangerous: mid-transaction data loss risk (no rollback on interrupt)

Step 5 - Write report

File lifetime-report.md in current working directory:

1. Summary - processes by classification.
2. Process table - most dangerous first. Per process: name, type, entry file, shutdown handler, grace period, recommendation.
3. Open questions.

Step 6 - Summarize

State report path, highlight processes that would lose data if terminated now.

Usage

/lifetime               # interactive
/lifetime <dir>         # scan project
/lifetime -help