rabbitmq-tactics
Fail
Audited by Snyk on Aug 12, 2026
Risk Level: CRITICAL
Full Analysis
HIGH W007: Insecure credential handling detected in skill instructions.
- Insecure credential handling detected (high risk: 1.00). The skill prompt explicitly shows and instructs embedding credentials and secrets (e.g., guest:guest, USER:PASS, literal JSON passwords, and STOLEN_COOKIE/Erlang cookie) directly into commands and payloads, which requires the LLM to handle/output secret values verbatim and is therefore insecure.
CRITICAL E005: Suspicious download URL detected in skill instructions.
- Suspicious download URL detected (high risk: 1.00). These URLs include direct download endpoints hosted on the target (http://TARGET:15672/cli/ and http://TARGET:15672/cli/rabbitmqadmin), which are untrusted sources for executable/script retrieval and thus represent a high-risk vector for malware distribution.
CRITICAL E006: Malicious code pattern detected in skill scripts.
- Malicious code pattern detected (high risk: 1.00). This content is an explicit offensive playbook instructing unauthorized access, credential abuse, data exfiltration (Shovel/Federation), creation of persistent backdoor accounts, and remote code execution via Erlang Cookie—clearly malicious in intent.
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 0.90). The skill includes a runtime wget that downloads and installs an executable from the target Management UI (http://TARGET:15672/cli/rabbitmqadmin), which fetches remote code that the skill then makes executable and runs, so it meets the criteria for a risky runtime external dependency.
MEDIUM W013: Attempt to modify system services in skill instructions.
- Attempt to modify system services in skill instructions detected (high risk: 1.00). This is an actionable exploitation playbook that instructs creating backdoor users, configuring persistent Shovel/Federation, reading Erlang cookies, and performing remote code execution and destructive actions—explicitly directing modification and persistence on the target system.
Issues (5)
W007
HIGHInsecure credential handling detected in skill instructions.
E005
CRITICALSuspicious download URL detected in skill instructions.
E006
CRITICALMalicious code pattern detected in skill scripts.
W012
MEDIUMUnverifiable external dependency detected (runtime URL that controls agent).
W013
MEDIUMAttempt to modify system services in skill instructions.
Audit Metadata