skills/mukul975/anthropic-cybersecurity-skills/building-soc-playbook-for-ransomware/Gen Agent Trust Hub
building-soc-playbook-for-ransomware
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTIONMETADATA_POISONING
Full Analysis
- [DATA_EXFILTRATION]: The check_id_ransomware function in scripts/agent.py enables the upload of a local file to an external domain (id-ransomware.malwarehunterteam.com). While intended for analyzing ransomware samples, this functionality could be abused to exfiltrate sensitive data if an attacker can manipulate the agent into designating a non-malicious file as a 'sample'. Additionally, the search_iocs_splunk function supports a SKIP_TLS_VERIFY environment variable that disables SSL/TLS certificate validation, which could expose session keys or search data to network interception.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for processing untrusted data found in ransomware samples and notes. Ingestion points: scripts/agent.py reads local files via the --sample argument for hashing and upload. Boundary markers: Absent; there are no instructions to the agent to sanitize or ignore instructions embedded within the files being processed. Capability inventory: The agent has capabilities for file system read access, outbound network communication, and interaction with sensitive enterprise platforms like Splunk and CrowdStrike. Sanitization: Absent; the content of the provided sample file is read and transmitted in its raw form.
- [METADATA_POISONING]: There is an inconsistency in the skill's identity metadata; the SKILL.md file lists the author as 'mahipal', whereas the LICENSE file and system context identify the author or copyright holder as 'mukul975'.
Audit Metadata