skills/mukul975/anthropic-cybersecurity-skills/performing-api-inventory-and-discovery/Gen Agent Trust Hub
performing-api-inventory-and-discovery
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The script scripts/agent.py executes the nslookup command via subprocess.run to perform DNS resolution. The call uses a list of arguments, which is a safe practice to prevent command injection.
- [EXTERNAL_DOWNLOADS]: The skill utilizes the requests library to fetch external resources, such as JavaScript files and Swagger specifications, from user-defined targets. In SKILL.md, some examples demonstrate disabling TLS verification (verify=False), which is a common but insecure practice in discovery tools.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, presenting a surface for indirect prompt injection. 1. Ingestion points: analyze_har_for_apis (SKILL.md) and parse_swagger_spec (scripts/agent.py) ingest data from HAR files and Swagger JSON specifications. 2. Boundary markers: The skill does not employ specific delimiters or instructions to ignore embedded commands. 3. Capability inventory: The skill possesses network communication (requests), shell command execution (nslookup), and cloud resource inventory (boto3) capabilities. 4. Sanitization: Ingested data is analyzed using standard libraries (json) and regular expressions.
Audit Metadata