The Agent Skills Directory

[DATA_EXFILTRATION]: The skill instructions facilitate the reading of sensitive credentials from ~/.gooseworks/credentials.json. The extracted api_key is then included in the Authorization header of network requests to an external API endpoint (api.gooseworks.ai).
[COMMAND_EXECUTION]: Setup procedures involve executing inline Python scripts (python3 -c) to parse JSON configuration files and export environment variables directly from the shell.
[EXTERNAL_DOWNLOADS]: The skill makes network calls to api.gooseworks.ai to proxy requests to the Didit service. This domain is not identified as a well-known or trusted service in the provided context.
[COMMAND_EXECUTION]: The skill references the execution of npx gooseworks login and npx goose-skills install, which involve fetching and running external code from the npm registry.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests untrusted user data and interpolates it into shell-based network commands without explicit sanitization.
Ingestion points: phone_number and code parameters provided by the user (interpolated in SKILL.md).
Boundary markers: Absent; the skill does not use delimiters or instructions to ignore embedded commands in the user input.
Capability inventory: Performs network operations via curl and local file reads via shell/Python commands.
Sanitization: Absent; the inputs are directly embedded into the request body of shell commands.

phone-verification