The Agent Skills Directory

REMOTE_CODE_EXECUTION (HIGH): The documentation explicitly recommends the command curl -sSL https://canifi.com/skills/google-search-console/install.sh | bash. Executing unverified code from an untrusted remote source via a pipe to a shell is a critical security risk that allows arbitrary system compromise.\n- EXTERNAL_DOWNLOADS (HIGH): The skill initiates downloads from canifi.com, which is not a trusted source. This applies to both the primary installation script and the canifi-env setup script.\n- CREDENTIALS_UNSAFE (HIGH): The skill prompts users to store high-value secrets such as GOOGLE_CLIENT_SECRET, GOOGLE_REFRESH_TOKEN, and SERVICE_PASSWORD using a local CLI tool. Given the unverifiable nature of the installation method, these credentials should be considered at high risk of exposure to the script author.\n- COMMAND_EXECUTION (MEDIUM): The skill relies on multiple shell operations (cp -r, canifi-env set) which could be abused if the underlying scripts are malicious or if local environment variables are compromised.\n- PROMPT_INJECTION (LOW): The skill is susceptible to indirect prompt injection via the data it processes from the Google Search Console API.\n
Ingestion points: Performance reports, top search queries, and URL inspection data from external Google APIs.\n
Boundary markers: None identified; untrusted data is likely interpolated directly into the agent's context.\n
Capability inventory: Subprocess execution (bash) and credential management via canifi-env.\n
Sanitization: No evidence of sanitization or validation for data returned by the search console.

google-search-console