error-propagation-mapper
Error propagation mapper: trace every error from origin through handling blocks to surface, LLM classifies each path as monitored/silent/dangerous. Read-only. Trigger: /error-map
SKILL.mdskills/runtime/error-propagation-mapper/SKILL.md
GitHub/error-map
Errors flow through code like water through pipes. This skill maps every path to find silent data loss and crash risks.
What this is for
- Errors swallowed in empty catch blocks
- Error context lost through wrapping without cause chaining
- Returned null defaults that callers treat as success
- Read-only skill. No code modification.
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 /error-map -help or /error-map -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/error-trace.ps1" -ProjectDir "<path>"
Step 4 - Classification
Read each error path:
- Monitored: error is logged and surfaced (HTTP error, alert)
- Silent: error is swallowed or returns default without notification
- Dangerous: error causes crash, data corruption, or info leak
Step 5 - Write report
File error-propagation-report.md in current working directory:
- Summary - error paths by classification.
- Path table - dangerous first. Per path: origin, handler chain, action taken, context preserved, final destination.
- Open questions.
Step 6 - Summarize
State report path, highlight dangerous silent paths.
Usage
/error-map # interactive
/error-map <dir> # scan project
/error-map -help