/sideorder

Systems fail not because individual operations fail but because they happen in the wrong order. This skill finds ordering risks.

What this is for

• Email sent before DB write confirms (user sees "confirmed" but order failed)
• Cache updated before DB write (stale cache after rollback)
• Irreversible operation before reversible one (charged credit card but order failed)
• Read-only skill. No code 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 /sideorder -help or /sideorder -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/sidechain-trace.ps1" -ProjectDir "<path>"

Step 4 - Classification

Read each handler's operation chain:

• Safe: transactional or order doesn't matter
• Risky: later failure leaves inconsistency
• Dangerous: data loss on crash mid-chain
• Inverted: non-atomic operation order could cause inconsistency

Step 5 - Write report

File side-effect-ordering-report.md in current working directory:

1. Summary - handlers by risk level.
2. Handler table - most dangerous first. Per handler: file, handler name, operation chain, ordering risk, recommendation.
3. Open questions.

Step 6 - Summarize

State report path, highlight handlers that would leave inconsistent state on failure.

Usage

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