job-posting-intent
Audited by Snyk on Jul 10, 2026
CRITICAL E006: Malicious code pattern detected in skill scripts.
- Malicious code pattern detected (high risk: 0.90). High-risk: the code embeds hard-coded bearer tokens and calls out to external services (rube.app / Gooseworks/Apify) including sending executable workbench code and scraped data, enabling data exfiltration and remote code-execution/backdoor abuse if those tokens or endpoints are abused.
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.75). The required runtime workflow calls an Apify LinkedIn Jobs scraper and then ingests outsider-authored job description text (e.g.,
descriptionText,company.description) into the script’s computed fields and exports them to the Google Sheet via Rube MCP, which is then used as LLM context for the agent’s “brief summary of top leads” (indirect prompt-injection risk from scraped public web content).
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 0.90). The scripts make runtime requests to Rube endpoints (https://rube.app/mcp and https://rube.app) and send code/commands (via RUBE_REMOTE_WORKBENCH / Composio tool calls) that are executed remotely, so these URLs are used at runtime to execute remote code.
HIGH W008: Secret detected in skill content (API keys, tokens, passwords).
- Secret detected (high risk: 1.00). I found a hardcoded, high-entropy JWT-style bearer token present verbatim in the code. The string (three base64 segments separated by dots) is not a placeholder like "YOUR_API_KEY" or "sk-xxxx" — it looks like a real access token and is repeated in multiple files. Hardcoding such tokens is a secret leakage risk and should be removed/rotated and replaced with environment variables.
Note: SKILL.md contains an example placeholder APIFY token (export APIFY_API_TOKEN="apify_api_YOUR_TOKEN_HERE") which is a documentation placeholder and is intentionally ignored.
Issues (4)
Malicious code pattern detected in skill scripts.
Third-party content exposure detected (indirect prompt injection risk).
Unverifiable external dependency detected (runtime URL that controls agent).
Secret detected in skill content (API keys, tokens, passwords).