The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references documentation and implementation guides hosted on an unofficial third-party domain (sosumi.ai) rather than the official Apple Developer portal. These links are used for core framework components, including SRSensorReader, SRSensor, and entitlement configurations in SKILL.md.
[DATA_EXFILTRATION]: The skill facilitates access to and processing of highly sensitive user information, including keyboard typing metrics (which can track sentiment and typing speed), speech audio/recognition results, and clinical health data such as ECG and PPG.
[INDIRECT_PROMPT_INJECTION]: The skill establishes an attack surface for indirect prompt injection due to the nature of the data it processes.
Ingestion points: Data is ingested from system sensors via SRSensorReader as described in SKILL.md and references/sensorkit-patterns.md.
Boundary markers: No delimiters or instructions to ignore embedded commands are present in the code snippets provided for processing keyboard or speech recognition data.
Capability inventory: The skill provides logic for fetching, iterating, and logging diverse user samples across multiple high-privilege sensors.
Sanitization: There is no evidence of data sanitization, filtering, or validation for transcribed speech or keyboard metrics before the data is processed or logged in the provided Swift examples.

sensorkit