The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted healthcare practitioner data from CSV files, Google Sheets, or direct user input. This data is then used to construct search queries and extraction parameters, creating a potential attack surface where malicious input could influence the agent's behavior.
Ingestion points: Processes provider lists from CSV files, Google Sheet URLs, and direct text input in SKILL.md.
Boundary markers: The skill uses structured extraction and parsing patterns, but does not define strict boundary markers or sanitization logic for the raw input data before it is processed by subsequent tools.
Capability inventory: The skill utilizes nimble search, nimble extract, and spawns sub-agents (nimble-researcher) with bypassPermissions to perform research.
Sanitization: Lacks specific sanitization steps for user-provided strings before they are used in sub-agent prompts or API calls.
[COMMAND_EXECUTION]: During the preflight and onboarding phase, the skill instructs the user to install the vendor's command-line interface globally using npm install -g @nimble-way/nimble-cli if it is not already present. While this is a standard setup procedure for this vendor's tools, it involves executing installation commands that modify the global environment.
[EXTERNAL_DOWNLOADS]: The skill downloads and installs the @nimble-way/nimble-cli package from the NPM registry. As this is the official package from the skill's author ('nimbleway'), it is a legitimate resource for the skill's operation.

healthcare-providers-enrich