skills/mukul975/anthropic-cybersecurity-skills/implementing-canary-tokens-for-network-intrusion/Gen Agent Trust Hub
implementing-canary-tokens-for-network-intrusion
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFE
Full Analysis
- [PRIVILEGE_ESCALATION]: The skill instructs the agent to modify system-level configuration files including /etc/hosts, /etc/ssh/ssh_config, and web server configurations in /etc/nginx/conf.d/. Modifying these paths typically requires elevated permissions.
- [INDIRECT_PROMPT_INJECTION]: The monitor action in scripts/agent.py reads from /var/log/canary_alerts.json. Because this log contains data generated by external entities (attackers triggering tokens), it serves as an ingestion point for untrusted data into the agent's context.
- Ingestion points: /var/log/canary_alerts.json (scripts/agent.py)
- Boundary markers: Absent
- Capability inventory: File write and network requests via requests (scripts/agent.py)
- Sanitization: None observed during log parsing
- [EXTERNAL_DOWNLOADS]: The agent interacts with external APIs at canarytokens.org and canary.tools to generate and manage security tokens. It also utilizes the canarytools Python library for enterprise console integration.
- [CREDENTIALS_UNSAFE]: The skill generates and plants simulated AWS credentials (Access Key ID and Secret Access Key) into sensitive file locations such as ~/.aws/credentials and .env files to serve as honeypots.
Audit Metadata