cold-email
Verdict safe, but 2 findings (2 medium), mainly data exfiltration. Most recent scan 2026-02-06.
The cold-email skill presents MEDIUM security risks primarily related to data handling and API key management. The skill transmits potentially sensitive lead data (PII, business intelligence) to a third-party service without explicit privacy disclosures or consent warnings. API key security guidance is insufficient, risking credential exposure. The skill lacks critical metadata (license, compatibility, allowed-tools) reducing transparency. Long-running sync requests (3-5 min) may cause agent timeouts and resource issues without proper async handling. Referenced files (lead.py, the.py) are missing from the package. No evidence of prompt injection, command injection, or malicious code execution was found. The skill appears functionally legitimate but requires improved security documentation and data handling transparency.
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
Generate hyper-personalized cold email sequences using AI. Turn lead data into high-converting outreach campaigns.
- Package
- 1 file, 5.5 KB
- SKILL.md
- 766 words, 12+ sections
- Code
- None. Instructions only.
SKILL.md covers: MachFive - AI Cold Email Generator · Setup · Campaign ID · Endpoints · List Campaigns (discover before generate) · Single Lead (Sync) · Batch (Async) · Lead Fields · Options · Errors · Usage Examples · Pricing
Findings
MEDIUM Hardcoded API Key Requirement Without Secure Storage Guidance Data Exfiltrationllm
The skill requires MACHFIVE_API_KEY environment variable but provides no guidance on secure storage, rotation, or scope limitations. Users may hardcode keys in shell profiles or insecure locations. The skill also accepts API keys via X-API-Key header, increasing exposure surface.
Setup instructions: 'Set `MACHFIVE_API_KEY` in your environment' and 'Or use `X-API-Key: {MACHFIVE_API_KEY}` header' without security warnings or best practices.MEDIUM Potential Data Exfiltration via Third-Party API Data Exfiltrationllm
The skill sends potentially sensitive lead data (names, titles, emails, LinkedIn URLs, company websites) to an external third-party service (app.machfive.io) without explicit user consent warnings or data handling disclosures. Users may unknowingly transmit PII and business intelligence to an external vendor.
POST requests to https://app.machfive.io/api/v1/campaigns/{campaign_id}/generate and /generate-batch send lead objects containing: name, title, company, email, company_website, linkedin_url. No privacy notice or data handling disclosure provided.What is in the package
| File | Kind | Size |
|---|---|---|
SKILL.md | doc | 5.5 KB |
External hosts referenced
app.machfive.io acme.com linkedin.com machfive.io
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
MACHFIVE_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/cold-email
- Raw scan result: machfive-cold-email-scan.json
- Badge for your README:
- Request a re-scan