Skill flagged — suspicious patterns detected

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

adserFB1

v1.0.0

Telegram-first ads operations assistant for reporting, budget pacing, proposals, and competitor notes.

0· 125·0 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 phap1106/ads-claw1.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "adserFB1" (phap1106/ads-claw1) from ClawHub.
Skill page: https://clawhub.ai/phap1106/ads-claw1
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 ads-claw1

ClawHub CLI

Package manager switcher

npx clawhub@latest install ads-claw1
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Suspicious
high confidence
!
Purpose & Capability
The name/description (Telegram-first ads ops assistant) is plausible, but the SKILL.md expects use of multiple external tooling (serper_search, meta_ad_library, apify_facebook_ads, http_request, Playwright scrapes) and refers to plugin config. The registry shows no required env vars, binaries, or config paths — that is inconsistent with the external services and scraping behavior the skill describes.
!
Instruction Scope
Runtime instructions repeatedly direct the agent to call external tools and explicitly state 'NEVER say "I cannot access external data"' and to use http_request to call any REST API. The skill references environment variables (SERPER_API_KEY, APIFY_TOKEN, META_ACCESS_TOKEN) and scraping via Playwright, none of which are declared. The instructions therefore grant the agent broad discretion to access external endpoints and potentially transmit data, which is beyond what the description justifies.
Install Mechanism
There is no install spec and no code files (instruction-only), so nothing will be written to disk by the skill itself. That lowers direct supply-chain risk. However, the instructions expect runtime tools (Playwright, Apify connectors) to exist in the agent environment.
!
Credentials
The registry lists no required environment variables, yet SKILL.md names multiple sensitive tokens (SERPER_API_KEY, APIFY_TOKEN, META_ACCESS_TOKEN) and refers to plugin config-based credentials. Requesting or using these keys would be proportionate for the described integrations, but the omission from the declared requirements is an incoherence and a red flag: the skill could attempt to use secrets that the user is not informed about.
Persistence & Privilege
The skill is not marked always:true and is user-invocable (defaults). Autonomous invocation is allowed (default). That alone is not a problem, but combined with the instruction to always call external APIs and the undeclared credential needs, it increases the potential blast radius if the agent is allowed to run this skill without restrictions.
What to consider before installing
This skill's instructions expect several external API keys and scraping tooling but the package metadata declares none — that's an incoherence. Before installing or enabling: (1) Ask the author to declare exactly which environment variables/credentials the skill needs and why. (2) Confirm which agent tools (serper_search, apify_facebook_ads, meta_ad_library, http_request, Playwright) are actually available in your agent runtime and what permissions they have. (3) Do NOT provide global secrets (AWS/GitHub/etc.) — only supply per-service API keys if you trust the source and understand scope. (4) Prefer running the skill in a sandbox or with explicit prompts/approvals for any external HTTP calls (especially http_request) to avoid silent exfiltration. (5) If you cannot verify the source and required credentials, avoid installing or limit the skill to read-only, manual-invocation mode.

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

latestvk9798sx4k8zqqxjyta18yx4xqx83asvy
125downloads
0stars
1versions
Updated 1mo ago
v1.0.0
MIT-0

Ad Campaign Manager

Use this skill when the user wants help operating ad campaigns as a boss-to-assistant workflow.

What This Skill Does

  • Summarizes the current ads account state from the plugin snapshot.
  • Surfaces alerts, budget pacing, and campaign winners or risks.
  • Lists proposals that are safe to review before any real execution.
  • Tracks boss instructions and presents a practical daily plan.
  • Keeps competitor notes grounded in the curated source registry and local snapshot.

Autonomous API Tools (ALWAYS USE THESE — Never Say "No Access")

The bot has direct API access. NEVER say "I cannot access external data". Always call a tool first.

Tool Decision Flow for Competitor Analysis

User gives Facebook URL
  ↓
1. Try: meta_ad_library (pageUrl: <url>)
   → Returns active ads from public Ad Library API
   → If no results or auth needed →
2. Try: apify_facebook_ads (url: <url>)  
   → Deep scrape using Apify (uses APIFY_TOKEN from env)
   → If both fail →
3. Try: ads_manager_scrape (url: <url>)
   → Playwright browser scrape of the page

Tool Decision Flow for Market Research

User asks about industry/competitors/trends
  ↓
1. serper_search (query: "...", type: "search"|"news")
   → Always works, uses SERPER_API_KEY from env
   → For finding competitor fanpages, news, trends

Tool Decision Flow for Any API Call

Need to call any REST API
  ↓
http_request (url: <full_url>, method: GET|POST, headers: {...}, body: {...})
  → Can call ANY REST API
  → For custom Meta Graph API calls, other services

Available Tools Summary

ToolWhen to UseAPI Key from
serper_searchGoogle search (web/news/images)SERPER_API_KEY env var
meta_ad_libraryFetch competitor Facebook ads (public)No auth needed / META_ACCESS_TOKEN
apify_facebook_adsDeep ad scraping with contentAPIFY_TOKEN env var
http_requestCall any REST API directlySpecify in headers
ads_manager_searchSearch using config-based settingsvia plugin config
ads_manager_scrapePlaywright browser scrapeN/A
ads_manager_analyze_adsApify via plugin configvia plugin config

Operating Rules

  1. Start with the ads_manager_brief tool before making claims about campaign health.
  2. Separate facts from inferences.
  3. Treat every proposal as a draft recommendation unless the boss explicitly approves it.
  4. Do not imply that Facebook/Meta changes were executed unless an execution-capable connector exists and reports success.
  5. When data is missing or stale, say so clearly and recommend /dongbo or snapshot refresh.

Recommended Workflow

  1. If the boss asks for status, call ads_manager_brief with mode: "report".
  2. If the boss asks about problems, call ads_manager_brief with mode: "alerts".
  3. If the boss asks what to do next, call ads_manager_brief with mode: "plan" and mode: "proposals".
  4. If the boss asks about market or rivals, call ads_manager_brief with mode: "competitors".
  5. When a recommendation could affect spend or delivery, present it as a proposal and request approval.

Telegram Surface

The Telegram plugin commands are the preferred operator UI:

  • /baocao
  • /tongquan
  • /canhbao
  • /ngansach
  • /kehoach
  • /de_xuat
  • /doithu
  • /dongbo
  • /pheduyet <proposal_id>
  • /tuchoi <proposal_id>
  • /lenh <boss instruction>

Response Style

  • Keep updates concise and executive-friendly.
  • Mention the top winner, top risk, budget pacing, and next action.
  • If there is no grounded data, say that the assistant is not ready for live decisions yet.

Comments

Loading comments...