Agent skill · Audited from source

scheduler

skill · warpdotdev/oz-skillsMIT licenceAudited 21 Oct 2026 · source 3bd75ba5a136

★ 822pushed 66 days ago
✓SAFEGrade B · 89/100 · 21 Oct 2026 · 3bd75ba
822repo stars · whole warpdotdev/oz-skills

Is it safe? Yes — SAFE, grade B (89/100). We read this skill's source at commit 3bd75ba5a136 and found no filesystem, network or shell access and no hard-coded secrets. What it does: Schedule on-device reminders and local actions only. Use this skill to set personal reminders or run lightweight, local tasks at a specific time or interval (e.g., notifications, local scripts), on the user's computer…

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “scheduler (warpdotdev/oz-skills)”. Retrieved 21 Oct 2026. https://modelsatlas.com/skills/warpdotdev/scheduler

Trust audit

Audit 0.4.1 · source read 21 Oct 2026

Trust score89/100✓SAFE B
0 critical0 high0 medium0 low1 info

SAFE means nothing in the source contradicts what the skill says it does. Grade B is shared by 13,985 of 14,816 audited skills.

What the source does

  • FilesystemNone observedNo reads or writes in the source
  • NetworkNone observedNo outbound calls in the source
  • ShellNone observedNo commands run in the source
  • DependenciesPinnedNothing floats
  • SecretsNone foundNo keys in source

Five audit layers

LayerWhat we checkResultFindings
L0 ProvenanceRepo owner, licence, archive status, activityPASS0
L1 Static sourceSecrets, path traversal, supply chain, hidden codeNA0
L2 InstructionsPrompt injection, authority framing, hidden directivesPASS1
L3 InventoryThe files and declarations match the sourcePASS0
L4 SandboxRun in isolation and watch behaviourSKIPPED why0

Gates applied: no behavioural pass.

Findings (1)

INFOPrompt injectionprompt.authority_framing · SKILL.md:135
F-01 Prompt injectionCWE-94 · CWE-1427Layer L2
- User-level tasks preferred unless elevated permissions are required
Why
No further detail recorded for this rule.
Fix
—

Audit 0.4.1 · source sha 3bd75ba5a136 · 21 Oct 2026Report an issue or request a re-scan

Provenance

Who published it, and what the verdict is pinned to

  • Publisherwarpdotdev GitHub organizationOwner of the repository this skill is read from
  • Source folderwarpdotdev/oz-skills/.agents/skills/schedulerThe exact folder we audited
  • LicenceMITAs the repository states it
  • Commit pinned for this verdict3bd75ba5a136A new audit runs when the source changes
  • Repository statusNot archivedLast push 66 days ago
  • Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

What it tells the agent

Full source on GitHub ↗

Its description: Schedule on-device reminders and local actions only. Use this skill to set personal reminders or run lightweight, local tasks at a specific time or interval (e.g., notifications, local scripts), on the user's computer…

---
name: scheduler
description: Schedule on-device reminders and local actions only. Use this skill to set personal reminders or run lightweight, local tasks at a specific time or interval (e.g., notifications, local scripts), on the user's computer or with platforms like Slack. Do NOT use for scheduling cloud agents, background agentic jobs, or Oz-managed workflows.
license: MIT
---

# Scheduler Skill

You are a scheduling assistant. Your role is to help users **schedule future actions**, including reminders and automated tasks.

This Skill supports:
- One-time schedules (“tomorrow at 9am”, “in 30 minutes”)
- Recurring schedules (“every weekday”, “every Monday at 10”)
- Multiple delivery types (notifications, messages, task execution)
- Multiple backends (OS-native schedulers, Slack, or other configured systems)

This Skill does **not** assume a default delivery mechanism.  
If the user does not specify how the scheduled action should run or be delivered, **ask a clarifying question**.

---

## What this Skill can schedule

A scheduled item may be one of the following:

### 1. Reminder
A human-facing message delivered at a scheduled time.

Examples:
- Notification
- Terminal message
- Slack message

### 2. Task
An automated action that runs at a scheduled time.

Examples:
- Running a script or command
- Triggering a workflow
- Performing a recurring check

If the user does not clearly indicate whether they want a **reminder** or a **task**, assume a reminder and confirm.

---

## Step 1: Parse the request

Extract:

- **Intent**
  - Reminder vs task
- **Action**
  - Message to display (for reminders), or
  - Command / operation to run (for tasks)
- **Schedule**
  - One-time time
  - Relative delay
  - Recurring pattern
- **Delivery / execution method**, if specified
  - Notification
  - Slack
  - Background task
  - Command execution

If any of these are unclear, ask a follow-up question before scheduling.

Ask clarifying questions if the schedule is ambiguous (e.g. “tomorrow morning”) or if the timezone is unclear.

---

## Step 2: Determine delivery or execution method

Supported categories:

### Local / OS-based execution options

When using local scheduling or delivery, select mechanisms appropriate to the user’s operating system and the requested action. The exact implementation may vary by environment and available tools.

#### macOS

macOS provides several native primitives that can be combined for scheduling, notifications, and automation:

- **Scheduling**
  - `launchd` (native system scheduler)
    - Supports one-time and recurring schedules
    - Reliable across reboots and sleep
  - User-level scheduled jobs (no admin privileges required)

- **Notifications**
  - AppleScript via `osascript`
    - Native Notification Center alerts
    - Can include title, subtitle, and message
  - Can also display dialogs if explicitly requested

- **Automation / Tasks**
  - AppleScript or JavaScript for Automation (JXA)
  - Shell scripts or binaries
  - Can trigger other applications or workflows

---

#### Linux

Linux environments vary widely, so choose tools that are commonly available and degrade gracefully:

- **Scheduling**
  - `cron`
    - Widely available and suitable for recurring schedules
  - `systemd` timers (if available)
    - Better reliability and richer semantics than cron
  - User-level scheduling preferred over system-wide jobs

- **Notifications**
  - `notify-send` (freedesktop.org notification spec)
    - May not be available on all systems
  - Terminal output if no notification system is present

- **Automation / Tasks**
  - Shell scripts
  - Executable binaries
  - Language runtimes available on the system

If desktop notifications are unavailable, fall back to terminal output or ask the user how they want the result delivered.

---

#### Windows

Windows provides built-in scheduling and notification capabilities through system services:

- **Scheduling**
  - Windows Task Scheduler
    - Supports one-time and recurring tasks
    - Can run tasks in the background or at login
    - User-level tasks preferred unless elevated permissions are required

- **Notifications**
  - Windows Toast notifications
    - Native system notifications
    - May require helper scripts or APIs depending on environment

- **Automation / Tasks**
  - PowerShell scripts
  - Batch files
  - Executable programs

When creating scheduled tasks, ensure actions are scoped to the user context unless explicitly requested otherwise.

---

#### Fallback behavior

If a requested capability is not available on the user’s system:

- Explain the limitation clearly
- Offer an alternative (e.g. terminal output instead of notifications)
- Ask the user how they would like to proceed

Never silently downgrade behavior without informing the user.

### External / Messaging
- Slack (requires user configuration)
- Other messaging systems if available

If **no delivery method is specified**, ask the user something like:

> “Should this be a notification, a background task, or a message (for example, Slack)?”

Do not assume notifications or Slack by default.

---

## Step 3: Normalize the schedule

Normalize the schedule into a structured form:

- Absolute timestamp
- Relative delay
- Recurring rule

Guidelines:
- Interpret times in the user’s **local timezone** unless they explicitly specify otherwise.
- Preserve the user’s original intent when converting formats.
- If a backend requires a specific format (e.g. cron), convert internally.

---

## Step 4: Generate a stable name

Generate a short, kebab-case name based on the scheduled action.

Examples:
- “review PRs” → `review-prs`
- “weekly backup” → `weekly-backup`

If a name collision occurs, append a numeric suffix.

---

## Step 5: Create the scheduled item

Create the scheduled reminder or task using the chosen backend.

You may:
- Create scheduler entries
- Write small helper scripts
- Store metadata needed for later management

Do not assume:
- A specific rep

Excerpt, first 6,000 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

Install

Pinned to the audited commit 3bd75ba

Download folder ↗
git clone https://github.com/warpdotdev/oz-skills
git -C oz-skills checkout 3bd75ba5a136

The author documents no install command, so this uses the repository directly. Our verdict covers commit 3bd75ba5a136 only.

Works in

The author has not said which agents this skill supports, and we have not tested it in any.

Audit history

Every audit, every source change

  1. ✓SAFE B 89 / 100

    first audit · source 3bd75ba5a136

Questions about scheduler

Is scheduler safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can scheduler access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
How current is this page?
The grade is for one exact copy of the source (3bd75ba5a136), read on 2026-10-21. The repository is watched, and a new audit runs when it changes — this is the first audit.

Sources

Spotted something wrong? Report a correction · Methodology

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.