The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill generates application code (KeyStoreServlet.java) that ingests untrusted data from HTTP request parameters. Ingestion points: request.getParameter('namespace') and request.getParameter('alias'). Capability inventory: CredStoreClient.sendRequest uses java.net.http.HttpClient to perform network operations against SAP Credential Store. Sanitization: The generated code only checks if parameters are blank but does not perform URL encoding or header value validation before interpolation into the sapcp-credstore-namespace header and the request URI.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted code from the user's project (Java source and XML files) to drive its transformation logic. Ingestion points: web.xml and Java files identified by Grep and Read tools. Boundary markers: The instructions do not provide delimiters or warnings to ignore embedded instructions in the source code being processed. Capability inventory: The skill utilizes Write, Edit, and Bash tools to modify the project and execute commands based on detected patterns.
[COMMAND_EXECUTION]: The skill provides and expects the execution of standard administrative and build commands including mvn clean install, cf env, and cf create-service. These are used for their intended purpose of verifying the migration and provisioning platform services.

keystore-credstore