notque/vexjoy-agent / skills/infrastructure/deploy
deploy
FlaggedDeploy and infrastructure: web deploy, cron, package install, shell configuration.
references/cron-automation.md
View on GitHub# Cron Job Auditor Skill
Static analysis of cron and scheduled job scripts against a 9-point reliability checklist. Produces structured PASS/FAIL/WARN results with severity classification (CRITICAL, HIGH, MEDIUM, LOW) and paste-ready code fixes for every finding. Audits are read-only and pattern-based -- scripts are never executed, because cron scripts may delete data, send emails, or modify production state.
## Reference Loading Table
| Signal | Load These Files | Why |
|---|---|---|
| Checking error handling, `set -e`, `pipefail`, `trap`, exit codes | `shell-error-handling.md` | Routes to the matching deep reference |
| Checking lock files, flock, PID files, concurrent execution | `concurrency-and-locks.md` | Routes to the matching deep reference |
| Checking logging, timestamps, log rotation, stderr routing | `logging-and-rotation.md` | Routes to the matching deep reference |
| Creating cron jobs, scheduled tasks, headless agents, wrapper scripts | `headless-cron-creator.md` | Routes to cron job creation methodology |
## Instructions
### Phase 1: DISCOVER
**Goal**: Locate all cron/scheduled scripts to audit.
**Step 1: Read repository CLAUDE.md** (if present) to understand project conventions before auditing.
**Step 2: Identify target scripts**
If the user provides specific paths, use those. Otherwise search these directories recursively:
```
scripts/*.sh, cron/*.sh, jobs/*.sh, bin/*.sh
```
Also check for scripts referenced in crontab files, Makefiles, or CI configs.
**Step 3: Validate targets**
For each discovered file:
- Confirm it exists and is readable
- Check it has a shell shebang (`#!/bin/bash`, `#!/bin/sh`, `#!/usr/bin/env bash`)
- Skip non-shell files (Python cron jobs, etc.) with a note -- this skill audits shell scripts only; it cannot replace shellcheck for syntax issues or analyze complex control flow beyond pattern matching
**Step 4: Log discovery results**
```markdown
## Scripts Found
1. scripts/daily_backup.sh (bash, 45 lines)
2. cron/cleanup.sh (bash, 22 lines)
3. jobs/sync_data.sh (SKIPPED: Python script)
```
**Gate**: At least one auditable shell script identified. Proceed only when gate passes.
### Phase 2: AUDIT
**Goal**: Run every check against every script. Run all 9 checks regardless of script size or apparent simplicity -- small scripts grow, and missing basics cause production incidents.
**Step 1: Read each script fully**
Read the entire file content. Do not sample or skip sections. If the script sources a common library file (`source ...` or `. ...`), read the sourced file too -- patterns provided by sourced libraries count as PASS (with a note indicating the source).
**Step 2: Run the 9-point checklist**
Use regex pattern matching for reliable, reproducible detection. Verify matches are not inside comments (`# ...`) before counting them -- when a match appears in a comment or string, note reduced confidence rather than silently accepting it.
| # | Check | Patterns | Severity |
|---|-------|----------|----------|
| 1 | Error handling | `set -e`, `set -o errexit`, `\|\| exit` | CRITICAL |
| 2 | Exit code checking | `$?`, `if [ $? -eq`, `&& ... \|\|` | HIGH |
| 3 | Logging with timestamps | `>> *.log`, `$(date)`, `date +` | HIGH |
| 4 | Log rotation | `find -mtime -delete`, `logrotate`, `tail -n` | MEDIUM |
| 5 | Working directory | `cd "$(dirname"`, `SCRIPT_DIR=`, absolute paths | HIGH |
| 6 | PATH environment | `PATH=`, `export PATH`, `source *env` | MEDIUM |
| 7 | Lock file / concurrency | `.lock`, `flock`, `.pid`, lock file check | HIGH |
| 8 | Cleanup on exit | `trap ... EXIT`, `trap ... cleanup`, `rm -rf *tmp` | MEDIUM |
| 9 | Failure notification | `mail -s`, `curl *webhook`, `notify`, `alert` | LOW |
For each check, record:
- PASS with line number where pattern found, OR
- FAIL/WARN with specific recommendation including a paste-ready code snippet (findings without fixes create work without guidance)
**Step 3: Calculate score**
```
Score = passed / total_checks * 100
```
Classify scripts: 90-100% Excellent, 70-89% Good, 50-69% Needs Work, <50% Critical.
**Gate**: All 9 checks run against every script. No checks skipped. Proceed only when gate passes.
### Phase 3: REPORT
**Goal**: Produce structured, actionable audit output. Do not modify any scripts -- report problems with recommendations only.
**Step 1: Format per-script results**
```
CRON JOB AUDIT: scripts/daily_backup.sh
==================================================
[PASS] Error handling (line 3)
[PASS] Logging with timestamps (line 12)
[FAIL] Lock file: No concurrent run prevention
[WARN] PATH environment: PATH not explicitly set
SCORE: 7/9 (78%) - Good
```
**Step 2: Provide recommendations**
Every FAIL and WARN must include a specific code snippet the user can paste. Keep recommendations proportional to the script's scope -- suggest lock files, not monitoring frameworks.
```bash
# Recommendation: Add lock file
LOCK_FILE="/tmp/daily_backup.lock"
exec 200>"$LOCK_FILE"
flock -n 200 || { echo "Already running"; exit 0; }
trap "rm -f $LOCK_FILE" EXIT
```
**Step 3: Produce aggregate summary**
If auditing multiple scripts:
```
AGGREGATE SUMMARY
=================
Scripts audited: 4
Average score: 72%
Critical issues: 2 (missing error handling)
Most common gap: Lock files (3/4 scripts missing)
```
**Gate**: Every finding has a recommendation. Report is complete. Audit is done.
## Error Handling
### Error: "No Shell Scripts Found"
Cause: Scripts in unexpected locations, or cron jobs written in Python/Ruby
Solution:
1. Ask user for explicit paths
2. Search broader: `**/*.sh` across the entire repository
3. Check crontab entries for referenced file paths
### Error: "Script Has No Shebang"
Cause: Script relies on default shell interpreter
Solution:
1. Still audit the script (treat as bash)
2. Add finding: "Missing shebang line" as MEDIUM severity
3. Recommend adding `#!/bin/bash` or `#!/usr/bin/env bash`
### Error: "Regex Produces False Positive"
Cause: Pattern matches in comments, strings, or unrelated context
Solution:
1. Verify match by reading surrounding lines for context
2. Check if match is inside a comment (`# ...`) and exclude
3. Report the finding but note reduced confidence
### Error: "Script Uses Non-Standard Patterns"
Cause: Custom error handling, logging frameworks, or wrapper functions
Solution:
1. Check if script sources a common library file
2. Read the sourced file for the missing patterns
3. If patterns exist in sourced libraries, mark as PASS with note
## Reference Loading
| Task type / signal | Load this reference |
|--------------------|---------------------|
| Checking error handling, `set -e`, `pipefail`, `trap`, exit codes | `references/shell-error-handling.md` |
| Checking lock files, flock, PID files, concurrent execution | `references/concurrency-and-locks.md` |
| Checking logging, timestamps, log rotation, stderr routing | `references/logging-and-rotation.md` |
| Creating cron jobs, scheduled tasks, headless agents, wrapper scripts | `references/headless-cron-creator.md` |
## References
### Best Practices Reference
```bash
#!/bin/bash
set -euo pipefail # Error handling
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
cd "$SCRIPT_DIR" # Working directory
PATH=/usr/local/bin:/usr/bin:/bin # Explicit PATH
LOCK="/tmp/$(basename "$0").lock" # Lock file
exec 200>"$LOCK"
flock -n 200 || { echo "Already running"; exit 0; }
LOG="logs/$(basename "$0" .sh)_$(date +%Y%m%d).log"
exec > >(tee -a "$LOG") 2>&1 # Logging
echo "$(date): Starting"
trap 'rm -f "$LOCK" /tmp/mytmp_*' EXIT # Cleanup
find logs -name "*.log" -mtime +30 -delete # Log rotation
# ... actual work ...
if [ $? -ne 0 ]; then # Failure notification
echo "FAILED" | mail -s "Cron Alert" admin@example.com
fi
```
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deployment | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |