swift-concurrency
Warn
Audited by Gen Agent Trust Hub on Jul 5, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Instruction #3 in
SKILL.mddirects the agent to fetch documentation fromhttps://sosumi.ai, an unofficial third-party documentation mirror. - [DATA_EXFILTRATION]: Instruction #2 in
SKILL.mddirects the agent to 'grep' through the local files of other installed skills. This encourages the agent to access and read data from other extensions, potentially bypassing intended isolation boundaries. - [PROMPT_INJECTION]: The skill instructs the agent to retrieve and process external Markdown from an untrusted source, which represents an indirect prompt injection attack surface.
- Ingestion points: External URLs from
sosumi.aidefined inSKILL.md. - Boundary markers: Absent; there are no instructions to use delimiters or ignore embedded commands in the fetched content.
- Capability inventory: The agent is instructed to use network retrieval and local file searching tools.
- Sanitization: No sanitization or validation of the fetched external content is performed.
Audit Metadata