The Agent Skills Directory

[PROMPT_INJECTION]: The skill ingests and processes full-text drug label information from the external Valyu API (api.valyu.ai). This data is presented to the agent, creating a surface for indirect prompt injection where malicious instructions could be embedded in the drug label content.
Ingestion points: scripts/search.mjs (external API response)
Boundary markers: Absent (data is returned as standard JSON fields)
Capability inventory: scripts/search (Bash and Node.js execution)
Sanitization: None detected for the API-returned content.
[COMMAND_EXECUTION]: The skill utilizes a Bash wrapper to execute its Node.js logic and provides instructions for the agent to use shell commands (e.g., find, node) to locate and run the scripts within the local filesystem.
[CREDENTIALS_UNSAFE]: The skill manages a Valyu API key by reading from and writing to a local configuration file at ~/.valyu/config.json. While this is standard practice for CLI-based tooling, it involves storing sensitive authentication tokens on the local disk.
[DATA_EXFILTRATION]: The skill performs network operations to api.valyu.ai to transmit search queries and the API key. While this is the intended and documented primary functionality of the skill, it involves sending user input and credentials to an external domain.

drug-labels-search