The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches trademark data from the official TMview public API.
Evidence: scripts/tmview-search.sh communicates with https://www.tmdn.org/tmview/api/search/results to retrieve trademark information.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes trademark names and applicant details from an external source, creating a potential surface for indirect prompt injection if an attacker were to register marks containing malicious instructions.
Ingestion points: Trademark records are retrieved from the tmdn.org API into the script scripts/tmview-search.sh.
Boundary markers: The script formats the output into a tabulated report with clear headers, which helps the agent distinguish search results from system instructions.
Capability inventory: The skill utilizes curl for network access and jq for parsing but does not execute external data as code.
Sanitization: The script uses jq to extract and normalize specific data fields (like Nice classes and status) before they are displayed to the agent.

trademark-search