@wonsukchoi/crondex 0.1.0

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 wonsukchoi
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package/README.md

@@ -0,0 +1,130 @@
+# crondex
+
+A public, growing directory of pre-made cron jobs — built so any AI agent
+(Claude, Codex, Hermes, OpenClaw, or a plain LLM with shell access) can pull
+one, tweak it, and schedule it. Clone it, grab a job, adjust the knobs, run it.
+
+## Quick start
+
+```bash
+git clone https://github.com/wonsukchoi/crondex.git
+cat crondex/catalog.json          # browse what's available
+cat crondex/jobs/devops/dependency-audit.yaml   # read one job
+```
+
+Give an agent the repo and a goal ("set up something that checks my repo
+health every morning") — it reads `catalog.json`, finds the closest match,
+adjusts `schedule`/`variables`, and wires it into whatever scheduler it has.
+
+## CLI
+
+Not yet published to the npm registry, so `npx crondex` doesn't resolve
+publicly yet — for now, run it from a clone:
+
+```bash
+git clone https://github.com/wonsukchoi/crondex.git && cd crondex
+node bin/crondex.js list                          # browse everything
+node bin/crondex.js list --category devops        # filter by category or --tag
+node bin/crondex.js show backup-reminder          # print one job's YAML
+node bin/crondex.js add backup-reminder --dest ./cron/backup-reminder.yaml
+```
+
+`add` copies the job's YAML as-is into your project — it's yours to edit
+from there, same as a `git clone` + copy, just one command. Once this is
+published, the same commands work via `npx crondex ...` with no clone
+needed.
+
+## Layout
+
+```
+crondex/
+├── bin/crondex.js         CLI: list / show / add
+├── catalog.json           generated index of every job — read this first
+├── schema/job.schema.json spec every job file follows
+├── jobs/
+│   ├── devops/
+│   └── productivity/
+└── scripts/
+    ├── build-catalog.js   regenerates catalog.json from jobs/**/*.yaml
+    └── validate-jobs.js   validates every job against the schema
+```
+
+## Available jobs
+
+| id | category | schedule | modes |
+|---|---|---|---|
+| `dependency-audit` | devops | `0 8 * * 1` | script + agent-prompt |
+| `log-cleanup` | devops | `30 3 * * *` | script |
+| `repo-health-check` | devops | `0 9 * * 1-5` | script + agent-prompt |
+| `backup-reminder` | devops | `0 9 * * *` | script |
+| `ssl-cert-expiry-check` | devops | `0 6 * * *` | script |
+| `uptime-ping-check` | devops | `*/15 * * * *` | script |
+| `cost-alert` | devops | `0 7 * * *` | script + agent-prompt |
+| `daily-standup-summary` | productivity | `0 8 * * 1-5` | script + agent-prompt |
+| `inbox-triage` | productivity | `0 7,13 * * 1-5` | agent-prompt only |
+| `weekly-report` | productivity | `0 16 * * 5` | script + agent-prompt |
+
+Full details (description, tags, variables) live in `catalog.json` and each
+job's YAML file.
+
+## How a job works
+
+Every job is one YAML file with a `runner`:
+
+- **`shell`** — run `command` directly. Zero LLM tokens, deterministic, no
+  judgment calls.
+- **`agent-prompt`** — hand the `prompt` field to an LLM agent each run.
+  Costs tokens, but can synthesize, prioritize, and draft prose.
+- **`hybrid`** — ships both. You (or your agent) pick per run: `command` to
+  save tokens and get raw data, or `prompt` when you want the LLM to
+  interpret it. `script_note` on the job explains exactly what you trade
+  away by choosing the script.
+
+`{{placeholders}}` in `command`/`prompt` resolve from `variables`. Each job
+also carries a `version` — bumped whenever `prompt`/`command` behavior
+changes, so if you've already scheduled a job elsewhere you can tell when
+the upstream copy has moved on without you.
+
+```yaml
+id: dependency-audit
+version: 1
+name: Dependency Vulnerability Audit
+description: ...
+category: devops
+tags: [security, dependencies]
+schedule: "0 8 * * 1"        # standard 5-field cron
+timezone: "UTC"
+runner: hybrid
+command: |
+  ...raw shell audit, zero tokens...
+prompt: |
+  ...instructions with {{repo_path}}, LLM synthesizes+prioritizes...
+script_note: what you lose by using `command` instead of `prompt`
+variables:
+  repo_path:
+    default: "."
+    description: ...
+compatible_agents: [claude, codex, hermes, openclaw, generic]
+```
+
+Full spec: [`schema/job.schema.json`](schema/job.schema.json).
+
+## Using a job
+
+1. Pick a job from `catalog.json` — check `modes` to see if it's script-only,
+   agent-prompt-only, or both.
+2. Decide script vs. agent-prompt if the job is `hybrid`: script saves
+   tokens, agent-prompt gives you more detail/judgment (see `script_note`).
+3. Override any `variables` and `schedule` for your case.
+4. Hand `command` or `prompt` plus `schedule` to your scheduler — system
+   crontab, a hosted cron, or your agent's own scheduling mechanism. This
+   repo defines *what* to run and *when*, not the executor.
+
+## Contributing a job
+
+See [`CONTRIBUTING.md`](CONTRIBUTING.md) — copy `templates/job.template.yaml`,
+fill it in, `npm run validate && npm run build-catalog`, open a PR.
+
+## License
+
+[MIT](LICENSE)

package/bin/crondex.js

@@ -0,0 +1,103 @@
+#!/usr/bin/env node
+// CLI over catalog.json — browse jobs, read one, or pull one into your project.
+import { readFileSync, writeFileSync, existsSync } from "node:fs";
+import { join } from "node:path";
+
+const ROOT = new URL("..", import.meta.url).pathname;
+const CATALOG = JSON.parse(readFileSync(join(ROOT, "catalog.json"), "utf8"));
+const PKG = JSON.parse(readFileSync(join(ROOT, "package.json"), "utf8"));
+
+const [, , cmd, ...args] = process.argv;
+
+function flag(name) {
+  const i = args.indexOf(`--${name}`);
+  return i === -1 ? undefined : args[i + 1];
+}
+
+function findJob(id) {
+  const meta = CATALOG.jobs.find((j) => j.id === id);
+  if (!meta) {
+    console.error(`no job named "${id}". Run "crondex list" to see options.`);
+    process.exit(1);
+  }
+  return meta;
+}
+
+function printHelp() {
+  console.log(`crondex — browse and pull pre-made cron jobs
+
+Usage:
+  crondex list [--category <name>] [--tag <name>]
+  crondex show <id>
+  crondex add <id> [--dest <path>]
+
+Examples:
+  crondex list --category devops
+  crondex show dependency-audit
+  crondex add backup-reminder --dest ./cron/backup-reminder.yaml
+`);
+}
+
+function catalogInfoLine() {
+  const viaNpx = process.env.npm_command === "exec";
+  const freshness = viaNpx
+    ? "running via npx — you're always on the latest catalog."
+    : "running from a local/global install — run `npm update -g crondex` (or re-pull) if new jobs seem missing.";
+  return `crondex v${PKG.version} — ${CATALOG.count} jobs. ${freshness}`;
+}
+
+function list() {
+  const category = flag("category");
+  const tag = flag("tag");
+  const jobs = CATALOG.jobs.filter(
+    (j) => (!category || j.category === category) && (!tag || j.tags.includes(tag))
+  );
+  console.log(catalogInfoLine());
+  console.log();
+  if (!jobs.length) {
+    console.log("no jobs matched.");
+    return;
+  }
+  for (const j of jobs) {
+    console.log(`${j.id}  [${j.category}]  (${j.modes.join(", ")})`);
+    console.log(`  ${j.description}`);
+    console.log();
+  }
+}
+
+function show(id) {
+  console.log(readFileSync(join(ROOT, findJob(id).path), "utf8"));
+}
+
+function add(id) {
+  const meta = findJob(id);
+  const dest = flag("dest") ?? `./${id}.yaml`;
+  if (existsSync(dest)) {
+    console.error(`${dest} already exists — refusing to overwrite. Pass --dest to choose another path.`);
+    process.exit(1);
+  }
+  writeFileSync(dest, readFileSync(join(ROOT, meta.path), "utf8"));
+  console.log(`wrote ${dest}`);
+}
+
+switch (cmd) {
+  case "list":
+    list();
+    break;
+  case "show":
+    if (!args[0]) {
+      console.error("usage: crondex show <id>");
+      process.exit(1);
+    }
+    show(args[0]);
+    break;
+  case "add":
+    if (!args[0]) {
+      console.error("usage: crondex add <id> [--dest <path>]");
+      process.exit(1);
+    }
+    add(args[0]);
+    break;
+  default:
+    printHelp();
+}

package/catalog.json

@@ -0,0 +1,263 @@
+{
+  "generated_by": "scripts/build-catalog.js",
+  "schema": "schema/job.schema.json",
+  "count": 10,
+  "jobs": [
+    {
+      "id": "backup-reminder",
+      "version": 1,
+      "name": "Backup Freshness Check",
+      "description": "Checks that your backup file actually got updated recently, and warns you if it's missing or stale. Use this if you have a backup process running somewhere and want to catch it silently failing before you need it.",
+      "category": "devops",
+      "tags": [
+        "backup",
+        "reliability",
+        "maintenance"
+      ],
+      "schedule": "0 9 * * *",
+      "timezone": "UTC",
+      "runner": "shell",
+      "modes": [
+        "script"
+      ],
+      "compatible_agents": [
+        "generic",
+        "claude",
+        "codex"
+      ],
+      "path": "jobs/devops/backup-reminder.yaml"
+    },
+    {
+      "id": "cost-alert",
+      "version": 1,
+      "name": "Cloud/API Cost Alert",
+      "description": "Checks your recent cloud or API spend against a budget and warns you if you're over or on pace to go over. Use this if you've ever been surprised by a bill.",
+      "category": "devops",
+      "tags": [
+        "cost",
+        "billing",
+        "budget"
+      ],
+      "schedule": "0 7 * * *",
+      "timezone": "UTC",
+      "runner": "hybrid",
+      "modes": [
+        "script",
+        "agent-prompt"
+      ],
+      "compatible_agents": [
+        "claude",
+        "codex",
+        "hermes",
+        "openclaw",
+        "generic"
+      ],
+      "path": "jobs/devops/cost-alert.yaml"
+    },
+    {
+      "id": "daily-standup-summary",
+      "version": 1,
+      "name": "Daily Standup Summary",
+      "description": "Pulls together what you did yesterday (commits, closed tasks) plus today's calendar, and drafts a quick 3-line standup update. Use this if you're tired of writing the same \"yesterday / today / blockers\" recap every morning.",
+      "category": "productivity",
+      "tags": [
+        "standup",
+        "reporting",
+        "daily"
+      ],
+      "schedule": "0 8 * * 1-5",
+      "timezone": "America/Los_Angeles",
+      "runner": "hybrid",
+      "modes": [
+        "script",
+        "agent-prompt"
+      ],
+      "compatible_agents": [
+        "claude",
+        "codex",
+        "hermes",
+        "openclaw",
+        "generic"
+      ],
+      "path": "jobs/productivity/daily-standup-summary.yaml"
+    },
+    {
+      "id": "dependency-audit",
+      "version": 1,
+      "name": "Dependency Vulnerability Audit",
+      "description": "Checks your project's dependencies for known security vulnerabilities and packages that are way out of date, then summarizes what to fix first. Use this if you want a heads-up before something exploitable slips into production.",
+      "category": "devops",
+      "tags": [
+        "security",
+        "dependencies",
+        "maintenance"
+      ],
+      "schedule": "0 8 * * 1",
+      "timezone": "UTC",
+      "runner": "hybrid",
+      "modes": [
+        "script",
+        "agent-prompt"
+      ],
+      "compatible_agents": [
+        "claude",
+        "codex",
+        "hermes",
+        "openclaw",
+        "generic"
+      ],
+      "path": "jobs/devops/dependency-audit.yaml"
+    },
+    {
+      "id": "inbox-triage",
+      "version": 1,
+      "name": "Inbox Triage",
+      "description": "Sorts your unread messages into \"needs a reply today\" vs. \"just FYI\" vs. \"spam,\" and drafts quick replies for the easy ones — it never sends anything without you approving it first. Use this if your inbox piles up and you want a head start each morning.",
+      "category": "productivity",
+      "tags": [
+        "email",
+        "triage",
+        "daily"
+      ],
+      "schedule": "0 7,13 * * 1-5",
+      "timezone": "America/Los_Angeles",
+      "runner": "agent-prompt",
+      "modes": [
+        "agent-prompt"
+      ],
+      "compatible_agents": [
+        "claude",
+        "codex",
+        "hermes",
+        "openclaw",
+        "generic"
+      ],
+      "path": "jobs/productivity/inbox-triage.yaml"
+    },
+    {
+      "id": "log-cleanup",
+      "version": 1,
+      "name": "Old Log File Cleanup",
+      "description": "Deletes old log files so they don't quietly fill up your disk. Use this if your app or server writes logs to a folder and nobody's manually clearing them out.",
+      "category": "devops",
+      "tags": [
+        "housekeeping",
+        "disk-space"
+      ],
+      "schedule": "30 3 * * *",
+      "timezone": "UTC",
+      "runner": "shell",
+      "modes": [
+        "script"
+      ],
+      "compatible_agents": [
+        "generic",
+        "claude",
+        "codex"
+      ],
+      "path": "jobs/devops/log-cleanup.yaml"
+    },
+    {
+      "id": "repo-health-check",
+      "version": 1,
+      "name": "Git Repo Health Check",
+      "description": "Checks a git repo for branches nobody's touched in a while, uncommitted changes sitting around, and pull requests waiting on review, then reports a short digest. Use this if you want a daily nudge on repo hygiene without checking everything by hand.",
+      "category": "devops",
+      "tags": [
+        "git",
+        "ci",
+        "maintenance"
+      ],
+      "schedule": "0 9 * * 1-5",
+      "timezone": "UTC",
+      "runner": "hybrid",
+      "modes": [
+        "script",
+        "agent-prompt"
+      ],
+      "compatible_agents": [
+        "claude",
+        "codex",
+        "hermes",
+        "openclaw",
+        "generic"
+      ],
+      "path": "jobs/devops/repo-health-check.yaml"
+    },
+    {
+      "id": "ssl-cert-expiry-check",
+      "version": 1,
+      "name": "SSL Certificate Expiry Check",
+      "description": "Checks how many days are left before a website's SSL certificate expires. Use this if you run a site and want a warning before the padlock breaks and visitors start seeing browser security errors.",
+      "category": "devops",
+      "tags": [
+        "ssl",
+        "tls",
+        "security",
+        "uptime"
+      ],
+      "schedule": "0 6 * * *",
+      "timezone": "UTC",
+      "runner": "shell",
+      "modes": [
+        "script"
+      ],
+      "compatible_agents": [
+        "generic",
+        "claude",
+        "codex"
+      ],
+      "path": "jobs/devops/ssl-cert-expiry-check.yaml"
+    },
+    {
+      "id": "uptime-ping-check",
+      "version": 1,
+      "name": "Uptime Ping Check",
+      "description": "Pings a URL on a schedule and flags it if it stops responding. Use this if you want a basic \"is my site down\" alert without paying for a monitoring service.",
+      "category": "devops",
+      "tags": [
+        "uptime",
+        "monitoring",
+        "reliability"
+      ],
+      "schedule": "*/15 * * * *",
+      "timezone": "UTC",
+      "runner": "shell",
+      "modes": [
+        "script"
+      ],
+      "compatible_agents": [
+        "generic",
+        "claude",
+        "codex"
+      ],
+      "path": "jobs/devops/uptime-ping-check.yaml"
+    },
+    {
+      "id": "weekly-report",
+      "version": 1,
+      "name": "Weekly Progress Report",
+      "description": "Rolls up a week's worth of commits and progress into a short report you can hand to your team. Use this if you dread writing the Friday status update from scratch every week.",
+      "category": "productivity",
+      "tags": [
+        "reporting",
+        "weekly"
+      ],
+      "schedule": "0 16 * * 5",
+      "timezone": "America/Los_Angeles",
+      "runner": "hybrid",
+      "modes": [
+        "script",
+        "agent-prompt"
+      ],
+      "compatible_agents": [
+        "claude",
+        "codex",
+        "hermes",
+        "openclaw",
+        "generic"
+      ],
+      "path": "jobs/productivity/weekly-report.yaml"
+    }
+  ]
+}

package/jobs/devops/backup-reminder.yaml

@@ -0,0 +1,26 @@
+id: backup-reminder
+version: 1
+name: Backup Freshness Check
+description: >
+  Checks that your backup file actually got updated recently, and warns you
+  if it's missing or stale. Use this if you have a backup process running
+  somewhere and want to catch it silently failing before you need it.
+category: devops
+tags: [backup, reliability, maintenance]
+schedule: "0 9 * * *"
+timezone: "UTC"
+runner: shell
+command: >
+  f={{backup_path}}; [ -e "$f" ] || { echo "MISSING: $f not found"; exit 1; };
+  age=$(( ( $(date +%s) - $(stat -f %m "$f" 2>/dev/null || stat -c %Y "$f") ) / 86400 ));
+  if [ "$age" -gt {{max_age_days}} ]; then echo "STALE: $f last updated $age days ago"; exit 1;
+  else echo "OK: $f last updated $age days ago"; fi
+variables:
+  backup_path:
+    default: "./backups/latest.tar.gz"
+    description: File or directory to check the mtime of.
+  max_age_days:
+    default: 1
+    description: Alert if the backup is older than this many days.
+compatible_agents: [generic, claude, codex]
+notes: Exits non-zero on missing/stale so a scheduler can alert on failure. Adjust stat flag for Linux vs macOS if running raw (command above handles both).

package/jobs/devops/cost-alert.yaml

@@ -0,0 +1,46 @@
+id: cost-alert
+version: 1
+name: Cloud/API Cost Alert
+description: >
+  Checks your recent cloud or API spend against a budget and warns you if
+  you're over or on pace to go over. Use this if you've ever been
+  surprised by a bill.
+category: devops
+tags: [cost, billing, budget]
+schedule: "0 7 * * *"
+timezone: "UTC"
+runner: hybrid
+command: >
+  if command -v aws >/dev/null 2>&1; then
+    start=$(date -d "-{{lookback_days}} days" +%Y-%m-%d 2>/dev/null || date -v-{{lookback_days}}d +%Y-%m-%d);
+    end=$(date +%Y-%m-%d);
+    aws ce get-cost-and-usage --time-period Start=$start,End=$end --granularity DAILY --metrics UnblendedCost 2>/dev/null || echo "aws cost explorer call failed (check credentials/permissions)";
+  elif command -v vercel >/dev/null 2>&1; then
+    vercel billing 2>/dev/null || echo "vercel CLI present but no billing subcommand available";
+  else
+    echo "no supported billing CLI found (aws/vercel) — use agent-prompt mode instead";
+  fi
+prompt: |
+  Check current spend for {{provider_hint}} covering the last {{lookback_days}} days.
+  1. Use whatever billing/cost API, CLI, or dashboard you have access to for this account.
+  2. Compare total spend to the {{budget_usd}} budget.
+  3. If over budget, or on pace to exceed it before the period ends, alert
+     with the overage amount and the top 3 cost drivers.
+  4. If under budget, report spend vs. budget in one line.
+script_note: >
+  `command` dumps raw provider CLI output (AWS/Vercel only) with no budget
+  comparison — parsing arbitrary billing JSON reliably in shell isn't
+  practical. `prompt` works with any provider your agent has access to and
+  does the budget comparison and driver breakdown.
+variables:
+  provider_hint:
+    default: "cloud provider"
+    description: Name the provider/account to check (e.g. "AWS", "Vercel", "OpenAI API").
+  budget_usd:
+    default: 100
+    description: Budget threshold in USD for the lookback window.
+  lookback_days:
+    default: 30
+    description: Rolling window of spend to evaluate.
+compatible_agents: [claude, codex, hermes, openclaw, generic]
+notes: Script mode only covers AWS/Vercel CLIs out of the box; extend `command` for other providers as needed.

package/jobs/devops/dependency-audit.yaml

@@ -0,0 +1,39 @@
+id: dependency-audit
+version: 1
+name: Dependency Vulnerability Audit
+description: >
+  Checks your project's dependencies for known security vulnerabilities and
+  packages that are way out of date, then summarizes what to fix first.
+  Use this if you want a heads-up before something exploitable slips into
+  production.
+category: devops
+tags: [security, dependencies, maintenance]
+schedule: "0 8 * * 1"
+timezone: "UTC"
+runner: hybrid
+command: >
+  cd {{repo_path}} &&
+  if [ -f package.json ]; then npm audit;
+  elif [ -f requirements.txt ]; then pip-audit -r requirements.txt;
+  elif [ -f Cargo.toml ]; then cargo audit;
+  else echo "no supported manifest (package.json/requirements.txt/Cargo.toml) found"; fi
+prompt: |
+  Run a dependency audit on the repo at {{repo_path}}.
+  1. Detect the package manager(s) in use (npm/pnpm/yarn/pip/cargo/etc).
+  2. Run each manager's native audit command (e.g. `npm audit`, `pip-audit`).
+  3. List vulnerabilities by severity, and packages more than {{staleness_days}}
+     days behind latest.
+  4. Summarize findings in under 200 words with a prioritized action list.
+  Do not modify any files — report only.
+script_note: >
+  `command` prints the raw audit tool output, zero tokens. `prompt` additionally
+  flags stale-but-not-vulnerable packages and ranks/prioritizes the findings for you.
+variables:
+  repo_path:
+    default: "."
+    description: Path to the repo to audit.
+  staleness_days:
+    default: 180
+    description: Flag packages whose latest release is older than this many days behind current.
+compatible_agents: [claude, codex, hermes, openclaw, generic]
+notes: Read-only job. Safe to run unattended.

package/jobs/devops/log-cleanup.yaml

@@ -0,0 +1,26 @@
+id: log-cleanup
+version: 1
+name: Old Log File Cleanup
+description: >
+  Deletes old log files so they don't quietly fill up your disk.
+  Use this if your app or server writes logs to a folder and nobody's
+  manually clearing them out.
+category: devops
+tags: [housekeeping, disk-space]
+schedule: "30 3 * * *"
+timezone: "UTC"
+runner: shell
+command: >
+  find {{log_dir}} -type f -name "{{log_pattern}}" -mtime +{{retention_days}} -delete
+variables:
+  log_dir:
+    default: "./logs"
+    description: Directory to clean.
+  log_pattern:
+    default: "*.log"
+    description: Glob pattern for files to consider.
+  retention_days:
+    default: 14
+    description: Delete files last modified more than this many days ago.
+compatible_agents: [generic, claude, codex]
+notes: Destructive job — dry-run with `find ... -print` instead of `-delete` before first use.

package/jobs/devops/repo-health-check.yaml

@@ -0,0 +1,42 @@
+id: repo-health-check
+version: 1
+name: Git Repo Health Check
+description: >
+  Checks a git repo for branches nobody's touched in a while, uncommitted
+  changes sitting around, and pull requests waiting on review, then reports
+  a short digest. Use this if you want a daily nudge on repo hygiene
+  without checking everything by hand.
+category: devops
+tags: [git, ci, maintenance]
+schedule: "0 9 * * 1-5"
+timezone: "UTC"
+runner: hybrid
+command: >
+  cd {{repo_path}} && now=$(date +%s) &&
+  echo "--- branches stale >{{stale_branch_days}}d ---" &&
+  git for-each-ref --format='%(committerdate:unix) %(refname:short)' refs/heads/ |
+    while read ts name; do age=$(( (now-ts)/86400 )); [ "$age" -gt {{stale_branch_days}} ] && echo "$name: ${age}d"; done;
+  echo "--- uncommitted changes ---"; git status --short;
+  command -v gh >/dev/null && { echo "--- open PRs ---"; gh pr list --state open; } || echo "(gh not installed, skipping PR check)"
+prompt: |
+  Check the health of the repo at {{repo_path}}.
+  1. List local/remote branches not touched in over {{stale_branch_days}} days.
+  2. Flag any uncommitted changes sitting for more than a day.
+  3. If a CI config exists, report the status of the last run on the default branch.
+  4. List open PRs older than {{stale_pr_days}} days with no recent activity.
+  Summarize as a short digest, most actionable item first.
+script_note: >
+  `command` prints raw stale-branch/status/PR-list data, zero tokens. `prompt`
+  additionally checks CI status and turns it into a prioritized digest.
+variables:
+  repo_path:
+    default: "."
+    description: Path to the repo to check.
+  stale_branch_days:
+    default: 30
+    description: Branches with no commits in this many days are flagged stale.
+  stale_pr_days:
+    default: 7
+    description: Open PRs with no activity in this many days are flagged.
+compatible_agents: [claude, codex, hermes, openclaw, generic]
+notes: Read-only job. Needs `gh` CLI or equivalent for PR/CI data if available.

package/jobs/devops/ssl-cert-expiry-check.yaml

@@ -0,0 +1,33 @@
+id: ssl-cert-expiry-check
+version: 1
+name: SSL Certificate Expiry Check
+description: >
+  Checks how many days are left before a website's SSL certificate expires.
+  Use this if you run a site and want a warning before the padlock breaks
+  and visitors start seeing browser security errors.
+category: devops
+tags: [ssl, tls, security, uptime]
+schedule: "0 6 * * *"
+timezone: "UTC"
+runner: shell
+command: >
+  host={{host}}; port={{port}};
+  exp=$(echo | openssl s_client -servername "$host" -connect "$host:$port" 2>/dev/null | openssl x509 -noout -enddate | cut -d= -f2);
+  [ -z "$exp" ] && { echo "ERROR: could not fetch certificate for $host:$port"; exit 1; };
+  exp_epoch=$(date -d "$exp" +%s 2>/dev/null || date -j -f "%b %d %T %Y %Z" "$exp" +%s);
+  now=$(date +%s);
+  days=$(( (exp_epoch-now)/86400 ));
+  if [ "$days" -lt {{warn_days}} ]; then echo "WARNING: $host cert expires in $days days ($exp)"; exit 1;
+  else echo "OK: $host cert expires in $days days ($exp)"; fi
+variables:
+  host:
+    default: "example.com"
+    description: Hostname to check.
+  port:
+    default: 443
+    description: Port the TLS service listens on.
+  warn_days:
+    default: 14
+    description: Alert if the cert expires within this many days.
+compatible_agents: [generic, claude, codex]
+notes: Zero-token, deterministic — no LLM mode needed for a plain expiry check.

package/jobs/devops/uptime-ping-check.yaml

@@ -0,0 +1,26 @@
+id: uptime-ping-check
+version: 1
+name: Uptime Ping Check
+description: >
+  Pings a URL on a schedule and flags it if it stops responding.
+  Use this if you want a basic "is my site down" alert without paying for
+  a monitoring service.
+category: devops
+tags: [uptime, monitoring, reliability]
+schedule: "*/15 * * * *"
+timezone: "UTC"
+runner: shell
+command: >
+  url={{url}}; timeout={{timeout_seconds}};
+  code=$(curl -o /dev/null -s -w '%{http_code}' --max-time "$timeout" "$url");
+  if [ "$code" -ge 200 ] && [ "$code" -lt 400 ]; then echo "OK: $url returned $code";
+  else echo "DOWN: $url returned $code"; exit 1; fi
+variables:
+  url:
+    default: "https://example.com"
+    description: URL to check.
+  timeout_seconds:
+    default: 10
+    description: Max seconds to wait for a response before treating it as down.
+compatible_agents: [generic, claude, codex]
+notes: Zero-token, deterministic. Run at a tighter interval (e.g. every 5-15 min) for real monitoring.

package/jobs/productivity/daily-standup-summary.yaml

@@ -0,0 +1,38 @@
+id: daily-standup-summary
+version: 1
+name: Daily Standup Summary
+description: >
+  Pulls together what you did yesterday (commits, closed tasks) plus
+  today's calendar, and drafts a quick 3-line standup update. Use this if
+  you're tired of writing the same "yesterday / today / blockers" recap
+  every morning.
+category: productivity
+tags: [standup, reporting, daily]
+schedule: "0 8 * * 1-5"
+timezone: "America/Los_Angeles"
+runner: hybrid
+command: >
+  cd {{repo_path}} && echo "--- commits last {{lookback_hours}}h ---" &&
+  git log --since="{{lookback_hours}} hours ago" --pretty=format:'%h %s'
+prompt: |
+  Draft a standup update for {{person_name}}.
+  1. Summarize commits and merged PRs from the last {{lookback_hours}} hours in
+     {{repo_path}}.
+  2. List today's calendar events if a calendar source is connected.
+  3. Ask the user (or infer from open tasks) what blockers to mention.
+  Output exactly 3 lines: Yesterday / Today / Blockers.
+script_note: >
+  `command` dumps the raw commit log, zero tokens. `prompt` additionally pulls
+  calendar events, infers blockers, and writes the 3-line Yesterday/Today/Blockers format.
+variables:
+  person_name:
+    default: "me"
+    description: Name to use in the summary.
+  repo_path:
+    default: "."
+    description: Repo to pull commit history from.
+  lookback_hours:
+    default: 24
+    description: Window of history to summarize.
+compatible_agents: [claude, codex, hermes, openclaw, generic]
+notes: Pairs well with a Slack/email delivery step if the agent has that integration.

package/jobs/productivity/inbox-triage.yaml

@@ -0,0 +1,32 @@
+id: inbox-triage
+version: 1
+name: Inbox Triage
+description: >
+  Sorts your unread messages into "needs a reply today" vs. "just FYI" vs.
+  "spam," and drafts quick replies for the easy ones — it never sends
+  anything without you approving it first. Use this if your inbox piles up
+  and you want a head start each morning.
+category: productivity
+tags: [email, triage, daily]
+schedule: "0 7,13 * * 1-5"
+timezone: "America/Los_Angeles"
+runner: agent-prompt
+prompt: |
+  Triage unread messages in {{inbox_source}} from the last {{lookback_hours}} hours.
+  1. Bucket each into: needs-reply-today, fyi-only, spam/newsletter.
+  2. For anything in needs-reply-today that looks routine (scheduling, simple
+     confirmations), draft a short reply but do NOT send — save as a draft.
+  3. Report counts per bucket and list the needs-reply-today items with a
+     one-line reason each.
+variables:
+  inbox_source:
+    default: "primary inbox"
+    description: Which inbox/connector to read from.
+  lookback_hours:
+    default: 18
+    description: Window of unread messages to consider.
+compatible_agents: [claude, codex, hermes, openclaw, generic]
+notes: >
+  Never auto-send. Draft-only by design — review before sending.
+  No script mode: categorization and reply drafting require judgment, and
+  inbox access varies too much by connector for a generic shell equivalent.

package/jobs/productivity/weekly-report.yaml

@@ -0,0 +1,39 @@
+id: weekly-report
+version: 1
+name: Weekly Progress Report
+description: >
+  Rolls up a week's worth of commits and progress into a short report you
+  can hand to your team. Use this if you dread writing the Friday status
+  update from scratch every week.
+category: productivity
+tags: [reporting, weekly]
+schedule: "0 16 * * 5"
+timezone: "America/Los_Angeles"
+runner: hybrid
+command: >
+  cd {{repo_path}} && echo "--- shortlog last {{lookback_days}}d ---" &&
+  git shortlog -sn --since="{{lookback_days}} days ago" &&
+  echo "--- commits ---" && git log --since="{{lookback_days}} days ago" --pretty=format:'%h %an %s'
+prompt: |
+  Build a weekly report for {{project_name}} covering the last {{lookback_days}} days.
+  1. Summarize merged PRs / commits in {{repo_path}}.
+  2. Summarize closed tasks/issues if a tracker is connected.
+  3. Note any metrics worth flagging (e.g. errors, usage) if a source is
+     connected.
+  4. Write it as 4 sections: Shipped, In Progress, Blocked, Next Week.
+  Keep the whole report under 300 words.
+script_note: >
+  `command` dumps raw commit/shortlog stats, zero tokens. `prompt` additionally
+  pulls tracker/metrics data and writes the narrative Shipped/In Progress/Blocked/Next Week report.
+variables:
+  project_name:
+    default: "this project"
+    description: Name to use in the report header.
+  repo_path:
+    default: "."
+    description: Repo to pull history from.
+  lookback_days:
+    default: 7
+    description: Rolling window the report covers.
+compatible_agents: [claude, codex, hermes, openclaw, generic]
+notes: Schedule for Friday afternoon by default; adjust `schedule` for your team's cadence.

package/package.json

@@ -0,0 +1,41 @@
+{
+  "name": "@wonsukchoi/crondex",
+  "version": "0.1.0",
+  "description": "A public directory of pre-made, agent-editable cron jobs.",
+  "type": "module",
+  "license": "MIT",
+  "repository": {
+    "type": "git",
+    "url": "git+https://github.com/wonsukchoi/crondex.git"
+  },
+  "homepage": "https://github.com/wonsukchoi/crondex",
+  "keywords": [
+    "cron",
+    "cron-jobs",
+    "ai-agent",
+    "claude",
+    "codex",
+    "automation",
+    "scheduler"
+  ],
+  "bin": {
+    "crondex": "bin/crondex.js"
+  },
+  "files": [
+    "bin",
+    "jobs",
+    "schema",
+    "catalog.json"
+  ],
+  "engines": {
+    "node": ">=18"
+  },
+  "scripts": {
+    "build-catalog": "node scripts/build-catalog.js",
+    "validate": "node scripts/validate-jobs.js"
+  },
+  "devDependencies": {
+    "ajv": "^8.17.1",
+    "js-yaml": "^4.1.0"
+  }
+}

package/schema/job.schema.json

@@ -0,0 +1,86 @@
+{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "$id": "https://github.com/wonsukchoi/crondex/schema/job.schema.json",
+  "title": "Crondex Job",
+  "description": "One pre-made cron job any AI agent (Claude, Codex, Hermes, OpenClaw, etc.) can pull, adjust, and schedule.",
+  "type": "object",
+  "required": ["id", "version", "name", "description", "schedule", "runner", "compatible_agents"],
+  "properties": {
+    "id": {
+      "type": "string",
+      "pattern": "^[a-z0-9]+(-[a-z0-9]+)*$",
+      "description": "Unique slug, matches filename without extension."
+    },
+    "version": {
+      "type": "integer",
+      "minimum": 1,
+      "description": "Bump when `prompt` or `command` behavior changes meaningfully, so agents that already scheduled this job know to re-check it. Do not bump for wording-only edits."
+    },
+    "name": { "type": "string" },
+    "description": { "type": "string" },
+    "category": { "type": "string" },
+    "tags": {
+      "type": "array",
+      "items": { "type": "string" }
+    },
+    "schedule": {
+      "type": "string",
+      "description": "Standard 5-field cron expression, e.g. '0 9 * * 1-5'."
+    },
+    "timezone": {
+      "type": "string",
+      "description": "IANA timezone, e.g. 'America/Los_Angeles'. Defaults to system local time if omitted."
+    },
+    "runner": {
+      "type": "string",
+      "enum": ["agent-prompt", "shell", "hybrid"],
+      "description": "'agent-prompt' hands `prompt` to an LLM agent each run (uses tokens, richer output). 'shell' executes `command` directly (zero tokens). 'hybrid' ships both — pick `command` to save tokens or `prompt` for more detail/judgment."
+    },
+    "prompt": {
+      "type": "string",
+      "description": "Task text given to the agent. Supports {{variable}} placeholders resolved from `variables`."
+    },
+    "command": {
+      "type": "string",
+      "description": "Shell command to run, for runner: shell or hybrid. Supports {{variable}} placeholders."
+    },
+    "script_note": {
+      "type": "string",
+      "description": "For hybrid jobs: what the zero-token `command` gives you vs. what only the LLM `prompt` mode adds (e.g. no narrative synthesis, no judgment calls)."
+    },
+    "variables": {
+      "type": "object",
+      "additionalProperties": {
+        "type": "object",
+        "required": ["default"],
+        "properties": {
+          "default": {},
+          "description": { "type": "string" }
+        }
+      },
+      "description": "User-adjustable knobs, keyed by placeholder name."
+    },
+    "compatible_agents": {
+      "type": "array",
+      "items": {
+        "type": "string",
+        "enum": ["claude", "codex", "hermes", "openclaw", "generic"]
+      }
+    },
+    "notes": { "type": "string" }
+  },
+  "allOf": [
+    {
+      "if": { "properties": { "runner": { "const": "agent-prompt" } } },
+      "then": { "required": ["prompt"] }
+    },
+    {
+      "if": { "properties": { "runner": { "const": "shell" } } },
+      "then": { "required": ["command"] }
+    },
+    {
+      "if": { "properties": { "runner": { "const": "hybrid" } } },
+      "then": { "required": ["prompt", "command"] }
+    }
+  ]
+}