apple-events
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The
references/fetch-worker.shandsync-from-kit.shscripts download content from the vendor's repository atgithub.com/FradSer/apple-sync-kit.gitusinggit cloneto facilitate backend deployment. - [COMMAND_EXECUTION]: The skill provides shell scripts (
deploy-systemd.sh,fetch-worker.sh,sync-from-kit.sh) for setup and deployment. These scripts perform operations such as cloning remote repositories and reading process environment data from/procto verify credential configuration. - [DATA_EXFILTRATION]: The instructions in
SKILL.mddirect the agent to search through Claude Code session transcripts stored at~/.claude/projects/to locate context for task continuation. This involves accessing local history files that may contain sensitive historical interaction data. - [PERSISTENCE]: The
deploy-systemd.shscript automates the creation of systemd user service drop-in configuration files in~/.config/systemd/user/to ensure that sync credentials and environment variables persist across sessions. - [INDIRECT_PROMPT_INJECTION]: The skill processes natural language arguments and data from reminders or calendar events, creating a potential surface for indirect prompt injection.
- Ingestion points: Natural language input via
$ARGUMENTSand data retrieved from external sources like EventKit or the SQLite sync database. - Boundary markers: No specific boundary markers or instructions to ignore embedded commands are specified in the prompts.
- Capability inventory: The skill uses the
eventCLI, which handles network communication and local file modifications. - Sanitization: No explicit sanitization or validation procedures for handling untrusted data are mentioned in the documentation.
Audit Metadata