recon-osint

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The prompt explicitly includes a curl that injects a HIBP API key header variable and asks the report to list "Leaked secrets" (if any), which would require the agent to accept and embed secret values verbatim in requests or outputs, creating exfiltration risk.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This content is an explicitly offensive OSINT/recon playbook — it provides step‑by‑step techniques to harvest emails/credentials (theHarvester, h8mail, HIBP), discover and download exposed cloud buckets and backups (aws s3 ls, Azure/GCP blob listing, wayback/interesting URLs), dork for secrets on GitHub, and identify exploitable targets (nmap, searchsploit, nuclei) to prioritize exploitation, and thus strongly facilitates unauthorized data exfiltration and system compromise (no backdoor/RCE payloads are embedded, but the guide directly enables malicious attacks).

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 1.00). The skill explicitly fetches and ingests public, untrusted content as part of its workflow (e.g., crt.sh certificate transparency queries, waybackurls/Wayback Machine, GitHub API/search, NVD API, Shodan, and public S3/Azure/GCP endpoints) in SKILL.md, and those results are parsed and used to drive further scanning and exploitation decisions — satisfying the criteria for potential indirect prompt injection.