The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill installs the things3-cli binary from an external GitHub repository (github.com/ossianhempel/things3-cli) using the go install command.\n- [COMMAND_EXECUTION]: The skill executes the things CLI tool to read the database and perform write operations via macOS URL schemes.\n- [COMMAND_EXECUTION]: The skill requires the user to grant "Full Disk Access" to the calling application, which is a high-privilege permission that allows access to sensitive system and user files.\n- [DATA_EXFILTRATION]: The skill reads data from the local Things 3 database, which contains personal task information, and exposes it to the agent's context.\n- [CREDENTIALS_UNSAFE]: The documentation encourages the use of THINGS_AUTH_TOKEN in the environment or via command-line flags. Passing tokens via CLI flags can expose them in process listings (e.g., via ps).\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests task content from the local database that could contain malicious instructions.\n
Ingestion points: things search, things inbox, things today, things upcoming.\n
Boundary markers: No boundary markers or instructions to ignore embedded content are specified.\n
Capability inventory: Shell command execution via the things CLI.\n
Sanitization: No input sanitization or validation is performed on the data retrieved from the database.

things-mac