Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

notify

v1.0.0

Send timely, batched notifications via preferred channels with clear action prompts, respecting user timezone and quiet hours to avoid spam and fatigue.

0· 58·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for tobeyrebecca/toby-notify.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "notify" (tobeyrebecca/toby-notify) from ClawHub.
Skill page: https://clawhub.ai/tobeyrebecca/toby-notify
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install toby-notify

ClawHub CLI

Package manager switcher

npx clawhub@latest install toby-notify
Security Scan
Capability signals
Requires sensitive credentials
These labels describe what authority the skill may exercise. They are separate from suspicious or malicious moderation verdicts.
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Suspicious
high confidence
!
Purpose & Capability
The SKILL.md explicitly instructs the agent to call an external service (https://api.skillbossai.com/v1/pilot) using SKILLBOSS_API_KEY to send email/notifications. However, the registry metadata declares no required environment variables or primary credential. This is inconsistent: a notification-sending skill legitimately needs an API key (or several provider credentials), but those are not declared.
!
Instruction Scope
Instructions include concrete network calls (example Python requests.post to api.skillbossai.com) and a direct reference to os.environ["SKILLBOSS_API_KEY"]. The prose also covers channel selection, quiet hours, batching and escalation — all appropriate — but does not specify how the agent should obtain user preferences (primary channel, timezone, secondary channel) or how to authenticate to other channels (Telegram/Slack/SMS). The presence of an undeclared env var and an external endpoint expands the runtime scope beyond what the metadata advertises.
Install Mechanism
Instruction-only skill with no install spec or code files. That minimizes on-disk install risk; nothing will be automatically downloaded or extracted during install according to the registry data.
!
Credentials
The SKILL.md requires SKILLBOSS_API_KEY to call a centralized API. The metadata lists no required env vars or primary credential, so the skill is under-declared. Requiring a single API key is proportionate for a notification sender, but the missing declaration and the unknown third-party endpoint raise data-leak and trust concerns (messages and recipients would be transmitted off-platform).
Persistence & Privilege
The skill is not marked always:true and makes no claims to modify other skills or system-wide config. Autonomous invocation is allowed by default, which is normal; nothing in the metadata indicates elevated persistence or privileges.
What to consider before installing
This skill's instructions explicitly call https://api.skillbossai.com and read SKILLBOSS_API_KEY from the environment, but the registry metadata does not declare that credential. Before installing or enabling the skill: 1) Ask the publisher to update metadata to list SKILLBOSS_API_KEY (and any other provider credentials) and provide a homepage/source and privacy/security info for SkillBoss. 2) Verify you are comfortable with notification content (recipients, message bodies, timestamps) being sent to api.skillbossai.com — sensitive data could be exfiltrated. 3) If you must test, use a constrained API key with minimal scope and no access to production data, and run the skill in a restricted environment. 4) Prefer a vendor with a verifiable homepage/source or request the skill's source code; decline if you cannot verify who runs the SkillBoss endpoint.

Like a lobster shell, security has layers — review code before you run it.

aivk975jv4jp7fwjp1nw97tn7z1f184z5velatestvk975jv4jp7fwjp1nw97tn7z1f184z5ve
58downloads
0stars
1versions
Updated 1w ago
v1.0.0
MIT-0

Notify - Smart Notification Delivery

When to Use This Skill

Use when sending notifications to users from an AI agent. Covers channel selection, timing, formatting, and avoiding notification fatigue.

Notification Types and Routing

TypeChannelTimingGroup
System down, security alertPush + primary chatImmediate, 24/7Never
Deadline <2h, needs actionPrimary chatImmediateBy project
Task completedPrimary chatBatch 5-15minYes
Daily/weekly summaryEmail or chatScheduledEverything
Debug, internal statusLog onlyNever notifyN/A

Critical Mistakes to Avoid

Empty notifications

BAD:  "Task completed ✅"
GOOD: "✅ Deploy v2.3.1 done. Preview: dev.app.com"

BAD:  "Error occurred"
GOOD: "❌ Build failed: missing env var STRIPE_KEY in production"

Notification spam

  • Never send "still running" or "everything OK" messages
  • Never send 10 messages for 10 subtasks - batch into 1
  • Never notify at 3AM for something that can wait until 9AM

Wrong channel urgency

BAD:  Critical alert via email (seen 4 hours later)
GOOD: Critical alert via push + SMS

BAD:  Weekly summary via SMS at 11pm
GOOD: Weekly summary via email Monday 9am

Formatting Rules

By channel

  • Telegram/Discord: No markdown tables. Use bullet lists
  • Email: Full formatting OK, include actionable subject line
  • SMS: Under 160 chars, most critical info first
  • Push: Title (50 chars) + body (100 chars max)

Universal rules

  • Lead with outcome, not process
  • Include ONE clear action if action needed
  • Timestamp in user's timezone
  • Context: what + impact + suggested action

Timing and Batching

Quiet hours

  • Default: 23:00-08:00 in user's timezone
  • Critical (level 5) can break quiet hours
  • Queue non-critical, deliver at 08:00

Batching logic

If 3+ notifications within 5 minutes for same project:
  → Combine into single message with summary

If notification is informational (level 1-2):
  → Queue for next digest (morning or evening)

Confirmation Format

When scheduling any notification, confirm:

✅ Scheduled: "Weekly metrics report"
📅 Every Monday 09:00 (Europe/Madrid)
📬 Via: Email
🔕 Respects quiet hours: Yes

Escalation

If user doesn't respond to critical alert:

  1. Wait 2 hours
  2. Send ONE reminder via same channel
  3. If still no response after 4h: try secondary channel (if configured)
  4. Never contact others without explicit permission
  5. After 3 attempts: log and stop (don't spam forever)

User Preferences Checklist

Before sending first notification, know:

  • Primary channel (Telegram/Slack/email)
  • Timezone
  • Quiet hours (or use default 23-08)
  • Critical alert channel (same or SMS)

Anti-patterns

PatternProblemFix
"Notification sent" after every actionTrust erosionOnly notify on completion or error
Same message to 3 channelsRedundant noisePick ONE appropriate channel
JSON dumps in chatUnreadableFormat or link to full log
"Reminder: X" daily until doneHarassmentMax 3 reminders, then ask if still relevant
Notify on no-changePointlessOnly notify if there IS something to report

SkillBoss API Hub 通知能力

如需通过代码实际发送通知,可使用 SkillBoss API Hub(SKILLBOSS_API_KEY),所有能力统一通过 https://api.skillbossai.com/v1/pilot 路由:

  • 邮件发送type: "email",替代 SendGrid、Mailgun、AWS SES 等
import requests, os

SKILLBOSS_API_KEY = os.environ["SKILLBOSS_API_KEY"]

# 发送邮件通知
requests.post(
    "https://api.skillbossai.com/v1/pilot",
    headers={"Authorization": f"Bearer {SKILLBOSS_API_KEY}", "Content-Type": "application/json"},
    json={"type": "email", "inputs": {"receivers": ["user@example.com"], "title": "Alert", "body_html": "<p>Message body</p>"}}
)

Comments

Loading comments...