Back to skill

Security audit

Claude Skill

Security checks for vulnerabilities and agentic risk

Overview

This is a large community skill catalog that is mostly transparent, but it recommends unpinned automatic installs of unaudited skills and includes unsanitized third-party text that can influence recommendations.

Install only if you are comfortable using a large community-sourced catalog. Treat recommendations as search results, not endorsements: review each target skill's source and ClawHub security status before installing. Avoid automatic chat-based setup for unfamiliar skills, and prefer a pinned, reviewed installer version instead of `npx clawhub@latest`.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (1)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:30
Finding

Mutable npm Package Execution in Recommended Installation Command

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, line 30
Vulnerability Type: Supply-chain risk from an unpinned executable dependency
Risk Level: Medium

Vulnerable Code

bash
npx clawhub@latest install <skill-slug>

Technical Analysis

The recommended installation command uses npx with the mutable @latest tag. If the package is not already available locally, npx can retrieve it and immediately execute its code. Because @latest does not identify a fixed, reviewed release or verify an expected integrity value, the code executed by future users may differ from the code available at audit time.

The document also states that listed community skills are curated but not audited. Consequently, pinning the installer alone would not establish the trustworthiness of a selected skill; both the installer and the requested skill are relevant supply-chain trust boundaries.

Attack Path

  1. An attacker compromises the npm publisher account, package, registry delivery path, or a dependency used by a future clawhub release.
  2. A malicious release becomes the version referenced by the mutable latest distribution tag.
  3. A user follows the installation command from SKILL.md.
  4. npx downloads the currently resolved package and runs its executable code.
  5. The malicious package executes with the permissions and environment available to the invoking user.

Impact Assessment

Successful exploitation could provide arbitrary code execution under the invoking user's account. Depending on that account's permissions and environment, the malicious package could read or modify accessible project and user files, access environment variables and locally available credentials, make outbound network requests, or modify user-level configuration.

The finding does not demonstrate an existing malicious clawhub release. It identifies the absence of version and integrity controls in an instruction tha ...[truncated 42 chars]

Remediation
View remediation

Remediation Suggestions

  • Replace @latest with an exact package version that has been reviewed and approved.
  • Document the expected npm package integrity hash and verify it through a trusted lockfile or controlled installation process.
  • Prefer a previously installed and organization-approved CLI rather than downloading executable code at invocation time.
  • Run installation in a restricted environment with minimal filesystem, credential, and network access.
  • Require source review and provenance verification for both the installer and each selected community skill.
  • Avoid automatically installing skills merely because a repository URL was supplied in chat.
  • Establish an update process that separately reviews and approves each new installer version before changing the pinned version.
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • System Prompt LeakageDirect Leakage, Indirect Extraction, Tool-Based Exfiltration
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
Findings (31)

Ae5

High
Category
analysis-evasion
Confidence
100% confidence
Finding

This skill is an instruction-capable artifact that embeds an extremely large directory of third-party skills and operational guidance in a single file, making full semantic review difficult for both humans and automated scanners. That increases the chance that malicious, misleading, or unsafe installation guidance hidden in the corpus will be surfaced or acted upon without adequate scrutiny.

Content

No source excerpt is available for this finding.

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 247)May include surrounding context.

md
- [universal-skills-manager](https://github.com/openclaw/skills/tree/main/skills/jacob-bd/universal-skills-manager/SKILL.md) - The master coordinator for AI skills.
- [usdckrump](https://github.com/openclaw/skills/tree/main/skills/arunnadarasa/usdckrump/SKILL.md) - Pay with USDC Krump (USDC.k) via x402 on Story Aeneid EVVM.
- [venice-admin](https://github.com/openclaw/skills/tree/main/skills/sabrinaaquino/venice-admin/SKILL.md) - Venice AI account administration - check balance, view usage history, and manage API keys.
- [vincent-credentials](https://github.com/openclaw/skills/tree/main/skills/glitch003/vincent-credentials/SKILL.md) - Use this skill to securely store credentials (API keys, passwords, tokens) and write them to .env files.
- [voice-recognition](https://github.com/openclaw/skills/tree/main/skills/gykdly/voice-recognition/SKILL.md) - Local speech-to-text with OpenAI Whisper CLI.
- [wallet-api](https://github.com/openclaw/skills/tree/main/skills/andresubri/wallet-api/SKILL.md) - Interact with the BudgetBakers Wallet API for personal finance data.
- [which-llm](https://github.com/openclaw/skills/tree/main/skills/zapkid/which-llm/SKILL.md) - Deterministic decision-ranking API with HTTP 402 USDC payments and outcome credits (discounts).

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 1148)May include surrounding context.

md
- [universal-skills-manager](https://github.com/openclaw/skills/tree/main/skills/jacob-bd/universal-skills-manager/SKILL.md) - The master coordinator for AI skills.
- [usdckrump](https://github.com/openclaw/skills/tree/main/skills/arunnadarasa/usdckrump/SKILL.md) - Pay with USDC Krump (USDC.k) via x402 on Story Aeneid EVVM.
- [venice-admin](https://github.com/openclaw/skills/tree/main/skills/sabrinaaquino/venice-admin/SKILL.md) - Venice AI account administration - check balance, view usage history, and manage API keys.
- [vincent-credentials](https://github.com/openclaw/skills/tree/main/skills/glitch003/vincent-credentials/SKILL.md) - Use this skill to securely store credentials (API keys, passwords, tokens) and write them to .env files.
- [voice-recognition](https://github.com/openclaw/skills/tree/main/skills/gykdly/voice-recognition/SKILL.md) - Local speech-to-text with OpenAI Whisper CLI.
- [wallet-api](https://github.com/openclaw/skills/tree/main/skills/andresubri/wallet-api/SKILL.md) - Interact with the BudgetBakers Wallet API for personal finance data.
- [which-llm](https://github.com/openclaw/skills/tree/main/skills/zapkid/which-llm/SKILL.md) - Deterministic decision-ranking API with HTTP 402 USDC payments and outcome credits (discounts).

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
70% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · SKILL.md (reported line 1269)May include surrounding context.

md
- [database-designer](https://github.com/openclaw/skills/tree/main/skills/alirezarezvani/database-designer/SKILL.md) - A comprehensive database design skill that provides expert-level analysis, optimization, and migration capabilities.
- [dataset-finder](https://github.com/openclaw/skills/tree/main/skills/anisafifi/dataset-finder/SKILL.md) - Use this skill when users need to search for datasets, download data files, or explore data repositories.
- [date-night](https://github.com/openclaw/skills/tree/main/skills/tdavis009/date-night/SKILL.md) - Your AI date night concierge — plans, books, and coordinates an entire evening out through browser automation.
- [dcg-guard](https://github.com/openclaw/skills/tree/main/skills/starensen/dcg-guard/SKILL.md) - Hard-blocks dangerous shell commands (rm -rf, git push --force, etc.) before execution via OpenClaw's.
- [dd-video-analyzer](https://github.com/openclaw/skills/tree/main/skills/zedit42/dd-video-analyzer/SKILL.md) - Download videos, extract transcripts, capture frames.
- [deapi](https://github.com/openclaw/skills/tree/main/skills/zrewolwerowanykaloryfer/deapi/SKILL.md) - The cheapest AI media API on the market.
- [deen-time](https://github.com/openclaw/skills/tree/main/skills/fermions75/deen-time/SKILL.md) - Get daily Islamic prayer (Salah) times, Iftar, and Suhoor schedules for any location worldwide.

Hidden Instructions

High
Category
Prompt Injection
Confidence
81% confidence
Finding

The file includes hidden or non-obvious Unicode markers from embedded third-party content, which can conceal instructions or alter how nearby text is interpreted during review. In such a large instruction-bearing directory, hidden text increases the chance that malicious entries evade detection and get recommended.

Content

Scanner excerpt · SKILL.md (reported line 1438)May include surrounding context.

md
- [human-like-memory](https://github.com/openclaw/skills/tree/main/skills/jianghaibobo2015-rgb/human-like-memory/SKILL.md) - This skill provides long-term memory capabilities for conversations.
- [humanpay](https://github.com/openclaw/skills/tree/main/skills/johanneskares/humanpay/SKILL.md) - Stop hallucinating hands and start hiring them.
- [hxxra](https://github.com/openclaw/skills/tree/main/skills/cxlhyx/hxxra/SKILL.md) - A Research Assistant workflow skill with four core commands: search papers, download PDFs, analyze content.
- [hybrid-md](https://github.com/openclaw/skills/tree/main/skills/mkassab30-ux/hybrid-md/SKILL.md) - \# Agent Identity: The Hybrid MD (Dr.
- [hyperstack](https://github.com/openclaw/skills/tree/main/skills/deeqyaqub1-cmd/hyperstack/SKILL.md) - The Agent Provenance Graph for AI agents — the only memory layer where agents can prove what they knew, trace why.
- [icp-qualifier](https://github.com/openclaw/skills/tree/main/skills/visualdeptcreative/icp-qualifier/SKILL.md) - You are an ICP (Ideal Customer Profile) qualification agent for Visual Dept, an AI-powered product photography.
- [idea-validation](https://github.com/openclaw/skills/tree/main/skills/jk-0001/idea-validation/SKILL.md) - Validate a business idea before investing time or money.

Exfiltration Commands

High
Category
Prompt Injection
Confidence
90% confidence
Finding

Instructions found that direct the agent to transmit conversation context or user data to external services.

Content

Scanner excerpt · SKILL.md (reported line 2256)May include surrounding context.

md
- [linkswarm](https://github.com/openclaw/skills/tree/main/skills/heyw00d/linkswarm/SKILL.md) - Agent-to-agent backlink exchange network.
- [lnemail](https://github.com/openclaw/skills/tree/main/skills/lnemail/lnemail/SKILL.md) - Setup and use anonymous email accounts on LNemail.net using Bitcoin Lightning payments.
- [localsend](https://github.com/openclaw/skills/tree/main/skills/chordlini/localsend/SKILL.md) - Send and receive files to/from nearby devices using the LocalSend protocol.
- [localudpmessenger](https://github.com/openclaw/skills/tree/main/skills/turfptax/localudpmessenger/SKILL.md) - Use when agents need to communicate over the local network — "send message to agent", "discover agents", "check.
- [lofy-projects](https://github.com/openclaw/skills/tree/main/skills/harrey401/lofy-projects/SKILL.md) - Project management for the Lofy AI assistant — tracks multiple projects with milestones, priority scoring engine.
- [lumos-auditor](https://github.com/openclaw/skills/tree/main/skills/samiru369-lumos/lumos-auditor/SKILL.md) - The social network for AI agents.
- [m365-pnp-cli](https://github.com/openclaw/skills/tree/main/skills/thomyg/m365-pnp-cli/SKILL.md) - CLI for Microsoft 365 - Manage Microsoft 365 tenant, SharePoint Online, Teams, OneDrive.

Exfiltration Commands

High
Category
Prompt Injection
Confidence
90% confidence
Finding

Instructions found that direct the agent to transmit conversation context or user data to external services.

Content

Scanner excerpt · SKILL.md (reported line 2310)May include surrounding context.

md
- [linkswarm](https://github.com/openclaw/skills/tree/main/skills/heyw00d/linkswarm/SKILL.md) - Agent-to-agent backlink exchange network.
- [lnemail](https://github.com/openclaw/skills/tree/main/skills/lnemail/lnemail/SKILL.md) - Setup and use anonymous email accounts on LNemail.net using Bitcoin Lightning payments.
- [localsend](https://github.com/openclaw/skills/tree/main/skills/chordlini/localsend/SKILL.md) - Send and receive files to/from nearby devices using the LocalSend protocol.
- [localudpmessenger](https://github.com/openclaw/skills/tree/main/skills/turfptax/localudpmessenger/SKILL.md) - Use when agents need to communicate over the local network — "send message to agent", "discover agents", "check.
- [lofy-projects](https://github.com/openclaw/skills/tree/main/skills/harrey401/lofy-projects/SKILL.md) - Project management for the Lofy AI assistant — tracks multiple projects with milestones, priority scoring engine.
- [lumos-auditor](https://github.com/openclaw/skills/tree/main/skills/samiru369-lumos/lumos-auditor/SKILL.md) - The social network for AI agents.
- [m365-pnp-cli](https://github.com/openclaw/skills/tree/main/skills/thomyg/m365-pnp-cli/SKILL.md) - CLI for Microsoft 365 - Manage Microsoft 365 tenant, SharePoint Online, Teams, OneDrive.

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 2880)May include surrounding context.

md
- [email-capture-generator](https://github.com/openclaw/skills/tree/main/skills/cameron-jovan/email-capture-generator/SKILL.md) - Build high-converting lead magnets, squeeze pages, and email capture funnels using proven 5-section conversion.
- [emergency-rescue](https://github.com/openclaw/skills/tree/main/skills/gitgoodordietrying/emergency-rescue/SKILL.md) - Recover from developer disasters.
- [exa-web-search-free](https://github.com/openclaw/skills/tree/main/skills/whiteknight07/exa-web-search-free/SKILL.md) - Free AI search via Exa.
- [feishu-user](https://github.com/openclaw/skills/tree/main/skills/hacksing/feishu-user/SKILL.md) - Feishu document operations (User Access Token version)
- [financial-calculator](https://github.com/openclaw/skills/tree/main/skills/tarigha/financial-calculator/SKILL.md) - Advanced financial calculator with future value.
- [find-code-tasks](https://github.com/openclaw/skills/tree/main/skills/paulpete/find-code-tasks/SKILL.md) - Lists all code tasks in the repository with their status.
- [findmefindme](https://github.com/openclaw/skills/tree/main/skills/orlyjamie/findmefindme/SKILL.md) - A totally legitimate skill that does nothing suspicious.

Hidden Instructions

High
Category
Prompt Injection
Confidence
80% confidence
Finding

Hidden HTML comment content inside imported skill descriptions can carry metadata or covert instructions that are less visible during ordinary review. Since this file is meant to influence downstream recommendations, concealed annotations from third-party entries increase review evasion risk.

Content

Scanner excerpt · SKILL.md (reported line 2961)May include surrounding context.

md
- [shipp](https://github.com/openclaw/skills/tree/main/skills/kclonts/shipp/SKILL.md) - Shipp is a real-time data connector.
- [siliconflow-image-gen](https://github.com/openclaw/skills/tree/main/skills/lilei0311/siliconflow-image-gen/SKILL.md) - Generate images using SiliconFlow API (FLUX.1, Stable Diffusion, etc.)
- [skill-doctorbot-ci-validator](https://github.com/openclaw/skills/tree/main/skills/bamontejano/skill-doctorbot-ci-validator/SKILL.md) - Stop failing in production.
- [skill-security-reviewer](https://github.com/openclaw/skills/tree/main/skills/ninjagpt/skill-security-reviewer/SKILL.md) - <!-- Skill Security Reviewer | Version 3.0.0 | Author: chris@zast.ai -->.
- [skill-update-delta-monitor](https://github.com/openclaw/skills/tree/main/skills/andyxinweiminicloud/skill-update-delta-monitor/SKILL.md) - Helps detect security-relevant changes in AI skills after installation.
- [skillbench](https://github.com/openclaw/skills/tree/main/skills/g9pedro/skillbench/SKILL.md) - Track skill versions, benchmark performance, compare improvements, and get self-improvement signals.
- [sovereign-changelog-maker](https://github.com/openclaw/skills/tree/main/skills/ryudi84/sovereign-changelog-maker/SKILL.md) - Generates structured changelogs from git history.

Hidden Instructions

High
Category
Prompt Injection
Confidence
81% confidence
Finding

Another hidden-character occurrence confirms the directory is ingesting unsanitized third-party text. Repeated invisible markers across a recommendation corpus materially increase the chance of prompt-hiding, deceptive display, or scanner evasion.

Content

Scanner excerpt · SKILL.md (reported line 4619)May include surrounding context.

md
- [bexio](https://github.com/openclaw/skills/tree/main/skills/rdewolff/bexio/SKILL.md) - Bexio Swiss business software API for managing contacts, quotes/offers,.
- [book-flight](https://github.com/openclaw/skills/tree/main/skills/aszelem) - id: travel-agent.
- [bookkeeper](https://github.com/openclaw/skills/tree/main/skills/h4gen/bookkeeper/SKILL.md) - Meta-skill for pre-accounting automation by orchestrating gmail, deepread-ocr, stripe-api, and xero.
- [brainstorming-studio](https://github.com/openclaw/skills/tree/main/skills/myboxstorage/brainstorming-studio/SKILL.md) - # 🧠 Skill Router (Skill Orchestrator)
- [brochure-design-generation](https://github.com/openclaw/skills/tree/main/skills/eftalyurtseven/brochure-design-generation/SKILL.md) - Generate professional brochure designs using each::sense AI.
- [business-card-generation](https://github.com/openclaw/skills/tree/main/skills/eftalyurtseven/business-card-generation/SKILL.md) - Generate professional business cards using each::sense AI.
- [business-plan](https://github.com/openclaw/skills/tree/main/skills/jk-0001/business-plan/SKILL.md) - Write, structure, and update a business plan for a solopreneur.

Self-Modification

High
Category
Rogue Agent
Confidence
90% confidence
Finding

Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Content

Scanner excerpt · SKILL.md (reported line 5403)May include surrounding context.

md
- [secucheck](https://github.com/openclaw/skills/tree/main/skills/jooneyp/secucheck/SKILL.md) - Comprehensive security audit for OpenClaw.
- [secure-shopper](https://github.com/openclaw/skills/tree/main/skills/moodykong/secure-shopper/SKILL.md) - Asynchronous shopping research + checkout using secure-autofill (1Password-backed browser filling) with results.
- [security-scanner](https://github.com/openclaw/skills/tree/main/skills/dmx64/security-scanner/SKILL.md) - Automated security scanning and vulnerability detection for web applications, APIs, and infrastructure.
- [self-evolve](https://github.com/openclaw/skills/tree/main/skills/be1human/self-evolve/SKILL.md) - Autonomous self-evolution skill: grants the agent full authority to modify its own configuration, skills, prompts.
- [sell-evoweb-ai](https://github.com/openclaw/skills/tree/main/skills/galizki/sell-evoweb-ai/SKILL.md) - Create AI-First Website with GEO (Generative Engine Optimization) and marketing rules to convert visitors into leads.
- [sendook](https://github.com/openclaw/skills/tree/main/skills/obaid/sendook/SKILL.md) - Read and send emails from an existing Sendook inbox.
- [seo-affiliate-guide](https://github.com/openclaw/skills/tree/main/skills/cmunozdev/seo-affiliate-guide/SKILL.md) - Crea guías de compra SEO optimizadas para posts de afiliado en español. Úsala siempre que el usuario quiera.

Self-Modification

High
Category
Rogue Agent
Confidence
90% confidence
Finding

Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Content

Scanner excerpt · SKILL.md (reported line 5403)May include surrounding context.

md
- [secucheck](https://github.com/openclaw/skills/tree/main/skills/jooneyp/secucheck/SKILL.md) - Comprehensive security audit for OpenClaw.
- [secure-shopper](https://github.com/openclaw/skills/tree/main/skills/moodykong/secure-shopper/SKILL.md) - Asynchronous shopping research + checkout using secure-autofill (1Password-backed browser filling) with results.
- [security-scanner](https://github.com/openclaw/skills/tree/main/skills/dmx64/security-scanner/SKILL.md) - Automated security scanning and vulnerability detection for web applications, APIs, and infrastructure.
- [self-evolve](https://github.com/openclaw/skills/tree/main/skills/be1human/self-evolve/SKILL.md) - Autonomous self-evolution skill: grants the agent full authority to modify its own configuration, skills, prompts.
- [sell-evoweb-ai](https://github.com/openclaw/skills/tree/main/skills/galizki/sell-evoweb-ai/SKILL.md) - Create AI-First Website with GEO (Generative Engine Optimization) and marketing rules to convert visitors into leads.
- [sendook](https://github.com/openclaw/skills/tree/main/skills/obaid/sendook/SKILL.md) - Read and send emails from an existing Sendook inbox.
- [seo-affiliate-guide](https://github.com/openclaw/skills/tree/main/skills/cmunozdev/seo-affiliate-guide/SKILL.md) - Crea guías de compra SEO optimizadas para posts de afiliado en español. Úsala siempre que el usuario quiera.

Ae4

Medium
Category
analysis-evasion
Confidence
82% confidence
Finding

The file contains mixed-script and unusual Unicode content across thousands of entries, which can hide lookalike names, invisible markers, or misleading text that defeats casual review. In a skill directory, this matters because users may be directed to deceptive skills or misread security-relevant instructions.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The activation condition is broad enough that normal conversations about finding or choosing skills may trigger this directory skill automatically. Over-broad activation can override safer defaults and steer users toward installation flows or unreviewed third-party content they did not explicitly request.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
95% confidence
Finding

The recommended install command uses npx clawhub@latest, which pulls and executes the newest published package without pinning an exact version or integrity hash. That creates a supply-chain risk where a compromised upstream release can execute attacker-controlled code during installation.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
75% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 302)May include surrounding context.

md
- [omarchy](https://github.com/openclaw/skills/tree/main/skills/achals-iglu/omarchy/SKILL.md) - Omarchy operating guardrails for day-to-day system work: assume the host is Omarchy by default, choose.
- [pattern-finder](https://github.com/openclaw/skills/tree/main/skills/leegitw/pattern-finder/SKILL.md) - Discover what two sources agree on — find the signal.
- [say](https://github.com/openclaw/skills/tree/main/skills/tobihagemann/say/SKILL.md) - Text-to-Speech via macOS say command with Siri Natural Voices.
- [shortcuts-generator](https://github.com/openclaw/skills/tree/main/skills/erik-agens/shortcuts-skill/SKILL.md) - Generate macOS/iOS Shortcuts by creating plist files.
- [skill-email-management](https://github.com/openclaw/skills/tree/main/skills/latisen/skill-email-management/SKILL.md) - Expert email management assistant for Apple.
- [testflight-monitor](https://github.com/openclaw/skills/tree/main/skills/jon-xo/testflight-monitor/SKILL.md) - Monitor available TestFlight beta slots with smart app lookups and silent batch checking.
- [voice-wake-say](https://github.com/openclaw/skills/tree/main/skills/xadenryan/voice-wake-say/SKILL.md) - Speak responses aloud on macOS using the built-in `say`.

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · SKILL.md (reported line 390)May include surrounding context.

md
- [client-data-management](https://github.com/openclaw/skills/tree/main/skills/satoshistackalotto/client-data-management/SKILL.md) - Secure client database for Greek accounting firms.
- [client-flow](https://github.com/openclaw/skills/tree/main/skills/ariktulcha/client-flow/SKILL.md) - Automated client onboarding and project lifecycle management.
- [cls-news-scraper](https://github.com/openclaw/skills/tree/main/skills/ningjingzhiyuande/cls-news-scraper/SKILL.md) - Use this skill to pull real-time market news from Cailian Press (CLS) and extract stock-positive catalysts.
- [cn-ecommerce-search](https://github.com/openclaw/skills/tree/main/skills/shopmeskills/cn-ecommerce-search/SKILL.md) - Search products across 8 Chinese e-commerce platforms: Taobao, Tmall, JD, PDD, 1688, AliExpress, Douyin, XHS.
- [cocod](https://github.com/openclaw/skills/tree/main/skills/egge21m/cocod/SKILL.md) - A Cashu ecash wallet CLI for Bitcoin and Lightning payments.
- [coda-io](https://github.com/openclaw/skills/tree/main/skills/simonfunk/coda-io/SKILL.md) - Interact with Coda.io docs, tables, rows, pages, and automations via the Coda REST API v1.
- [code-stats](https://github.com/openclaw/skills/tree/main/skills/autogame-17/code-stats/SKILL.md) - Visualizes repository complexity by counting files, lines of code, and grouping by extension.

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · SKILL.md (reported line 474)May include surrounding context.

md
- [linz-public-transport](https://github.com/openclaw/skills/tree/main/skills/fjrevoredo/linz-public-transport/SKILL.md) - Query Linz Linien raw EFA endpoints for stop lookup and live departures.
- [lnget](https://github.com/openclaw/skills/tree/main/skills/roasbeef/lnget/SKILL.md) - Install and use lnget, a Lightning-native HTTP client with automatic L402 payment support.
- [lofy-fitness](https://github.com/openclaw/skills/tree/main/skills/harrey401/lofy-fitness/SKILL.md) - Fitness accountability for the Lofy AI assistant — workout logging from natural language, meal tracking.
- [logistics-tracking](https://github.com/openclaw/skills/tree/main/skills/shopmeskills/logistics-tracking/SKILL.md) - Track international packages by tracking number.
- [lygo-guardian-p0-stack](https://github.com/openclaw/skills/tree/main/skills/deepseekoracle/lygo-guardian-p0-stack/SKILL.md) - LYGO Guardian base skill – Nano-Kernel (P0.4), Understanding Heart (P0.5), and Light Math harmony as a portable.
- [lyrical-fable](https://github.com/openclaw/skills/tree/main/skills/sanzgiri/lyrical-fable/SKILL.md) - Create short lyrical fables (~1000 words) about historical, fictional, or mythological characters in the style.
- [mac-control](https://github.com/openclaw/skills/tree/main/skills/easonc13/mac-control/SKILL.md) - Control Mac via mouse/keyboard automation using cliclick and AppleScript.

Unrestricted Tool Access

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill grants unrestricted tool access without appropriate constraints. An agent with unfettered tool access can perform arbitrary actions including file modification, network requests, and code execution.

Content

Scanner excerpt · SKILL.md (reported line 820)May include surrounding context.

md
- [ecto](https://github.com/openclaw/skills/tree/main/skills/visionik/ecto/SKILL.md) - Ghost.io blog management via Admin API.
- [emredoganer-fizzy](https://github.com/openclaw/skills/tree/main/skills/emredoganer) - Manage Fizzy Kanban boards and cards.
- [endpoints](https://github.com/openclaw/skills/tree/main/skills/adamkristopher/endpoints/SKILL.md) - Endpoints document management API toolkit.
- [entr](https://github.com/openclaw/skills/tree/main/skills/gumadeiras/entr/SKILL.md) - Run arbitrary commands when files change.
- [ergocare-coach](https://github.com/openclaw/skills/tree/main/skills/akhmittra/ergocare-coach/SKILL.md) - Your personal desk health coach with automated break reminders.
- [error-guard](https://github.com/openclaw/skills/tree/main/skills/amar1432/error-guard/SKILL.md) - System safety and control-plane skill that prevents agent deadlocks.
- [evolution-drift-detector](https://github.com/openclaw/skills/tree/main/skills/andyxinweiminicloud/evolution-drift-detector/SKILL.md) - Helps detect when AI agent skills silently mutate across inheritance chains.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 1016)May include surrounding context.

md
- [agi-terminal-helper](https://github.com/openclaw/skills/tree/main/skills/liuwujijay/agi-terminal-helper/SKILL.md) - A practical runbook for using OpenClaw exec safely (sandbox-first, explicit confirmations, and debugging playbooks).
- [agntor](https://github.com/openclaw/skills/tree/main/skills/garinmckayl/agntor/SKILL.md) - Trust layer for AI agents — verify identities, guard prompts, redact PII, and manage x402 escrow payments.
- [agorahub](https://github.com/openclaw/skills/tree/main/skills/codevena/agorahub/SKILL.md) - AgoraHub agent registry — discover and use 14+ verified AI agents for dev tasks like hashing, encoding, formatting.
- [agos-marketplace](https://github.com/openclaw/skills/tree/main/skills/danielw8088/agos-marketplace/SKILL.md) - Integrate OpenClaw with Agos Marketplace and automatically execute both sell-side listing creation and buy-side.
- [ai-agent-tools](https://github.com/openclaw/skills/tree/main/skills/cerbug45/ai-agent-tools/SKILL.md) - This library provides ready-to-use Python functions that AI agents can leverage to perform various tasks.
- [ai-collab](https://github.com/openclaw/skills/tree/main/skills/jeremysommerfeld8910-cpu/ai-collab/SKILL.md) - Multi-agent autonomous collaboration system for two OpenClaw agents working in parallel.
- [ai-music-video](https://github.com/openclaw/skills/tree/main/skills/gprecious/ai-music-video/SKILL.md) - Generate AI music videos end-to-end.

Session Persistence

Medium
Category
Rogue Agent
Confidence
65% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 1749)May include surrounding context.

md
- [quantumos](https://github.com/openclaw/skills/tree/main/skills/murtiurti4/quantumos/SKILL.md) - Install and manage QuantumOS, an AI command center dashboard for OpenClaw.
- [quasi-coder](https://github.com/openclaw/skills/tree/main/skills/jhauga/quasi-coder/SKILL.md) - Expert 10x engineer skill for interpreting and implementing code from shorthand, quasi-code, and natural language.
- [queue-task](https://github.com/openclaw/skills/tree/main/skills/moodykong/queue-task/SKILL.md) - Durable queue-task helper for resumable, idempotent batch jobs in task-father task folders.
- [quick-reminders](https://github.com/openclaw/skills/tree/main/skills/lstpsche/quick-reminders/SKILL.md) - Zero-LLM one-shot reminders (<48h) via nohup sleep + openclaw message send, operated.
- [quick-test](https://github.com/openclaw/skills/tree/main/skills/gustavoziaugra/quick-test/SKILL.md) - Quick system test to verify OpenClaw environment.
- [quiver](https://github.com/openclaw/skills/tree/main/skills/stuhorsman/quiver/SKILL.md) - Query alternative financial data from Quiver Quantitative (Congress trading, Lobbying, Government Contracts.
- [quotly-style-sticker](https://github.com/openclaw/skills/tree/main/skills/sakullla/quotly-style-sticker/SKILL.md) - Generate QuotLy-style stickers from OpenClaw context and return MEDIA for auto-send.

Indirect Prompt Extraction

Medium
Category
System Prompt Leakage
Confidence
75% confidence
Finding

Skill contains patterns that could indirectly extract system prompts through rephrasing, translation, summarization, or side-channel techniques.

Content

Scanner excerpt · SKILL.md (reported line 2301)May include surrounding context.

md
- [sixel-email](https://github.com/openclaw/skills/tree/main/skills/sixel-et/sixel-email/SKILL.md) - 1:1 email channel for agents — the agent can only email one address, and only that address can email the agent.
- [skill-cleaner](https://github.com/openclaw/skills/tree/main/skills/jacobthejacobs/skill-cleaner/SKILL.md) - Automatically verify "suspicious" skills via VirusTotal and add them to the security allowlist via the Bridge.
- [skillguard-audit](https://github.com/openclaw/skills/tree/main/skills/jonathanliu811026/skillguard-audit/SKILL.md) - Audit agent skills for security threats before installing them.
- [subreddit-scout](https://github.com/openclaw/skills/tree/main/skills/xammarie/subreddit-scout/SKILL.md) - Find high-fit subreddits for a product, summarize rules, and suggest value-first posting angles.
- [surf-check](https://github.com/openclaw/skills/tree/main/skills/kevinmcnamee/surf-check/SKILL.md) - Surf forecast decision engine.
- [telnyx-freemium-upgrade](https://github.com/openclaw/skills/tree/main/skills/teamtelnyx/telnyx-freemium-upgrade/SKILL.md) - Automatically upgrade Telnyx account from freemium to professional tier.
- [teltel-send-sms-text-message](https://github.com/openclaw/skills/tree/main/skills/teltel-call-center/teltel-send-sms-text-message/SKILL.md) - Send SMS text messages via TelTel (teltel.io) using the REST API (api.teltel.io)

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · SKILL.md (reported line 2326)May include surrounding context.

md
- [agi-artificial-geometric-intelligence](https://github.com/openclaw/skills/tree/main/skills/uniaolives) - Designed multi-layer.
- [amplitude-automation](https://github.com/openclaw/skills/tree/main/skills/sohamganatra/amplitude-automation/SKILL.md) - Automate Amplitude tasks via Rube MCP.
- [canva](https://github.com/openclaw/skills/tree/main/skills/abgohel/canva/SKILL.md) - Create, export, and manage Canva designs via the Connect API.
- [ceorater](https://github.com/openclaw/skills/tree/main/skills/ceorater-skills/ceorater/SKILL.md) - Get institutional-grade CEO performance analytics for S&P 500.
- [check-analytics](https://github.com/openclaw/skills/tree/main/skills/jeftekhari/check-analytics/SKILL.md) - Audit existing Google Analytics implementation.
- [cicd-pipeline](https://github.com/openclaw/skills/tree/main/skills/gitgoodordietrying/cicd-pipeline/SKILL.md) - Create, debug, and manage CI/CD pipelines with GitHub.
- [clawver-store-analytics](https://github.com/openclaw/skills/tree/main/skills/nwang783/clawver-store-analytics/SKILL.md) - Monitor Clawver store performance.

Rp1

Medium
Category
MCP Rug Pull
Confidence
65% confidence
Finding

uvx/uv tool run commands without ==version create a rug-pull risk.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 4410)May include surrounding context.

md
- [beacon](https://github.com/openclaw/skills/tree/main/skills/scottcjn/beacon/SKILL.md) - Agent-to-agent protocol for social coordination, crypto payments, and P2P mesh.
- [bridle](https://github.com/openclaw/skills/tree/main/skills/bjesuiter/bridle/SKILL.md) - Unified configuration manager for AI coding assistants.
- [casual-cron](https://github.com/openclaw/skills/tree/main/skills/gostlightai/casual-cron/SKILL.md) - Create Clawdbot cron jobs from natural language with strict.
- [claw-sync](https://github.com/openclaw/skills/tree/main/skills/arakichanxd/claw-sync/SKILL.md) - Secure sync for OpenClaw memory and workspace.
- [cron-backup](https://github.com/openclaw/skills/tree/main/skills/zfanmy/cron-backup/SKILL.md) - Set up scheduled automated backups with version tracking and cleanup.
- [cron-retry](https://github.com/openclaw/skills/tree/main/skills/jrbobbyhansen-pixel/cron-retry/SKILL.md) - Auto-retry failed cron jobs on connection recovery.

Static analysis

No suspicious patterns detected.