The Agent Skills Directory

[DATA_EXFILTRATION]: The skill accesses a hardcoded sensitive file path in the user's home directory to retrieve API credentials.
Evidence: In scripts/cf-api.sh, the variable CF_CREDS_FILE is set to "$HOME/cloudflare_global_key". The script uses cat to read this file and extract the API token or global key.
Context: Hardcoding access to credentials in the root of the home directory is a high-risk pattern for data exposure.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data from the Cloudflare API and reflects it to the agent without security boundaries.
Ingestion points: Multiple scripts including scripts/dns.sh, scripts/zones.sh, and scripts/analytics.sh fetch data from the Cloudflare API endpoints.
Boundary markers: The skill does not use delimiters or instructions to the agent to ignore potentially malicious content within the API responses.
Capability inventory: The skill possesses high-privilege capabilities including the ability to delete DNS records, delete zones, and deploy Cloudflare Worker scripts.
Sanitization: No sanitization is performed on the data returned by the API (such as DNS record names or firewall descriptions) before it is outputted to the agent's context.
[COMMAND_EXECUTION]: The skill relies on shell commands to perform its core functions, including network communication and data processing.
Evidence: Scripts utilize curl for API interaction, jq for JSON manipulation, and standard utilities like cat, grep, and sed for credential extraction and file handling.

cloudflare