ABM Outbound
6 findings (2 critical, 2 high, 2 medium), mainly data exfiltration, social engineering, unauthorized tool use. Most recent scan 2026-02-05.
CRITICAL SECURITY RISK - This skill implements a comprehensive personal data collection and mass outreach pipeline with severe privacy, legal, and ethical violations. The skill orchestrates bulk scraping of LinkedIn profiles, enrichment with personal contact information (emails, phone numbers), skip tracing to obtain physical home addresses, and multi-channel outreach campaigns. Critical issues include: (1) Mass PII collection without consent mechanisms, (2) Insecure credential handling via environment variables, (3) Data exfiltration to multiple third-party services without adequate disclosure, (4) Potential violations of GDPR, CCPA, CAN-SPAM, TCPA, and LinkedIn ToS, (5) Ethical concerns with skip tracing and unsolicited physical mail to home addresses, (6) Misleading description that omits legal and privacy risks, (7) No data protection, retention, or deletion policies. This skill should NOT be deployed without comprehensive legal review, privacy impact assessment, consent mechanisms, and significant architectural changes to address data protection requirements. The skill poses significant legal liability risks to users and their organizations.
Scanner's overall assessment (LLM analyzer)
Second scanner: NVIDIA SkillSpector
Independent static run, no language model, over the same files as the Cisco scan above. How the two scanners compare across the whole corpus: Two scanners, one corpus.
What this skill does
Multi-channel ABM automation that turns LinkedIn URLs into coordinated outbound campaigns. Scrapes profiles, enriches with Apollo (email + phone), gets mailing addresses via Skip Trace, then orchestrates email sequences, LinkedIn touches, and handwritten letters via Scribeless. The secret weapon for standing out in crowded inboxes.
- Package
- 3 files, 13 KB
- SKILL.md
- 734 words, 12+ sections
- Code
- None. Instructions only.
SKILL.md covers: ABM Outbound · Prerequisites · Pipeline · Step 1: Gather LinkedIn URLs · Step 2: Scrape LinkedIn Profiles · Step 3: Enrich with Apollo (Email & Phone) · Step 4: Get Mailing Address (Skip Trace) · Step 5: Multi-Channel Outreach · 5a: Email Sequence · 5b: LinkedIn Sequence · 5c: Handwritten Letter (Scribeless) · Coordinated Timing
Findings
CRITICAL Hardcoded API Keys Required in Environment Variables Data Exfiltrationllm
The skill requires users to export sensitive API keys (APIFY_API_KEY, APOLLO_API_KEY, SCRIBELESS_API_KEY) as environment variables. This creates multiple security risks: (1) API keys stored in shell history files, (2) Keys accessible to all processes running under the user account, (3) No secure credential management, (4) Keys may be logged or exposed in process listings. The skill provides no guidance on secure credential storage or rotation.
SKILL.md
```bash
export APIFY_API_KEY="your_key"
export APOLLO_API_KEY="your_key"
export SCRIBELESS_API_KEY="your_key"
```CRITICAL Mass Personal Data Collection and Exfiltration Pipeline Data Exfiltrationllm
The skill orchestrates a comprehensive personal data collection pipeline that scrapes LinkedIn profiles, enriches with contact information (emails, phone numbers), obtains physical mailing addresses via skip tracing, and transmits all data to multiple third-party services (Apify, Apollo, Scribeless). This creates severe privacy and data protection risks: (1) Bulk collection of personal identifiable information (PII) without explicit consent, (2) Cross-service data sharing with multiple vendors, (3) Physical address lookup via skip tracing services, (4) No data retention or deletion policies, (5) Potential violations of GDPR, CCPA, and other privacy regulations, (6) No user consent mechanisms for data subjects.
SKILL.md
Pipeline stages: LinkedIn scraping → Email/phone enrichment (Apollo) → Physical address lookup (Skip Trace) → Multi-channel outreach (email, LinkedIn, physical mail). Example: 'curl -X POST "https://api.apify.com/v2/acts/one-api~skip-trace/run-sync-get-dataset-items"' for obtaining physical addresses.HIGH Misleading Skill Description Omits Privacy and Legal Risks Social Engineeringllm
The skill description presents the tool as a 'secret weapon for standing out in crowded inboxes' without disclosing the significant privacy, legal, and ethical risks involved in mass personal data collection, skip tracing, and unsolicited multi-channel outreach. The description fails to mention: (1) Potential violations of privacy laws (GDPR, CCPA), (2) Anti-spam regulations (CAN-SPAM, CASL), (3) Telephone Consumer Protection Act (TCPA) compliance for phone outreach, (4) Ethical concerns with skip tracing and physical mail to home addresses, (5) LinkedIn Terms of Service violations for scraping. This constitutes social engineering by omission, misleading users about the true nature and risks of the tool.
SKILL.md
Description: 'Multi-channel ABM automation that turns LinkedIn URLs into coordinated outbound campaigns... The secret weapon for standing out in crowded inboxes.' No mention of legal compliance, privacy risks, or consent requirements.HIGH Unauthorized Third-Party Service Integration Without Disclosure Unauthorized Tool Usellm
The skill integrates with multiple third-party services (Apify, Apollo, Scribeless, Instantly) and transmits user data to these services without adequate disclosure of: (1) Data sharing agreements, (2) Third-party data retention policies, (3) Vendor security practices, (4) Data processor agreements required under GDPR, (5) Liability for third-party data breaches. Users are instructed to sign up for these services and provide API keys without understanding the full data sharing implications.
SKILL.md
Prerequisites table lists 4 external services with sign-up links but no disclosure of data sharing, privacy policies, or vendor security practices.MEDIUM No Rate Limiting or Resource Consumption Controls Resource Abusellm
The skill provides no guidance on rate limiting, batch size limits, or resource consumption controls when processing prospect lists. Users could inadvertently trigger: (1) API rate limit violations resulting in account suspension, (2) Excessive API costs from bulk operations, (3) Service disruptions from overwhelming third-party APIs, (4) Denial of service to the user's own systems from unbounded operations. The curl examples show synchronous API calls without retry logic, timeout handling, or batch processing guidance.
SKILL.md
Curl examples use 'run-sync-get-dataset-items' endpoints without timeout parameters, batch size limits, or rate limiting guidance. No mention of maximum list sizes or processing limits.MEDIUM Promotes Potentially Unethical Cold Outreach Practices Social Engineeringllm
The skill promotes aggressive multi-channel outreach including unsolicited physical mail to home addresses obtained via skip tracing. This raises ethical concerns: (1) Skip tracing to obtain home addresses for commercial outreach, (2) Multi-channel bombardment (email + LinkedIn + physical mail), (3) No opt-out mechanisms mentioned, (4) Potential harassment through coordinated multi-channel campaigns, (5) Invasion of privacy by sending commercial mail to personal residences. The skill frames these practices as a 'secret weapon' without addressing ethical boundaries or recipient consent.
SKILL.md
Description promotes 'handwritten letters via Scribeless' using addresses from 'Skip Trace' as a competitive advantage. Step 4 shows obtaining physical addresses: 'curl -X POST "https://api.apify.com/v2/acts/one-api~skip-trace/run-sync-get-dataset-items"'What is in the package
| File | Kind | Size |
|---|---|---|
SKILL.md | doc | 7.4 KB |
references/enrichment.md | doc | 2.2 KB |
references/scribeless-api.md | doc | 3.6 KB |
External hosts referenced
platform.scribeless.co linkedin.com api.apify.com api.apollo.io api.instantly.ai apify.com apollo.io docs.scribeless.co instantly.ai
Hostnames that appear in the skill's files. Referencing a host is not a finding; compare against the findings above and the skill's stated purpose.
Environment variables referenced
APIFY_API_KEY APOLLO_API_KEY INSTANTLY_API_KEY SCRIBELESS_API_KEY
Package contents recorded 2026-10-22. The scan above ran against the version current at scan time; the two can differ if the skill has since been updated.
What we checked
- Static analyzer. Pattern rules over SKILL.md and bundled files: known-bad commands, exfiltration URLs, secrets, obfuscation.
- Behavioral analyzer. Checks what the skill asks the agent to do against what its description claims it does.
- Llm analyzer. A language model reads the skill and reports intent-level issues the rules cannot express.
- Trigger analyzer. Looks for instructions that fire on conditions the user did not ask for.
- Meta analyzer. A second model pass that validates, deduplicates and re-scores the other analyzers' findings.
What we did not check
- Runtime behaviour. The skill was not executed; findings come from reading its files.
- Anything fetched at install or run time, including remote scripts, package dependencies, and URLs the skill tells the agent to open.
- Content the analyzers cannot read: images, binaries, and some non-English text.
- Known scanner bypass classes. See the methodology and limitations page.
Sources
- Repository: https://clawhub.ai/skills/abm-outbound
- Raw scan result: abm-outbound-scan.json
- Badge for your README:
- Request a re-scan