rebound
Pass
Audited by Gen Agent Trust Hub on Apr 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local shell scripts (
rebound-cli.sh,rebound-relay) and Gradle tasks (reboundSnapshot,reboundSummary) to interface with the Android/iOS development environment and fetch live performance metrics.\n- [EXTERNAL_DOWNLOADS]: The skill requires the installation of theio.github.aldefy.reboundGradle plugin and associated compiler artifacts from standard package repositories as part of its setup guide.\n- [DATA_EXFILTRATION]: The skill implements an optional telemetry system (ReboundTelemetry.enabled) that, when enabled by the user, transmits anonymized aggregate performance statistics to the author's infrastructure.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests data from external snapshots that could be influenced by the application being monitored.\n - Ingestion points: JSON snapshot data containing composable function names, parameter states, and parent hierarchy strings (referenced in
references/cli-usage.md).\n - Boundary markers: Absent; there are no instructions for the agent to use delimiters or ignore embedded directives within the ingested JSON content.\n
- Capability inventory: The skill allows execution of subprocesses and shell commands to interact with the project and devices (referenced in
SKILL.mdandreferences/setup-guide.md).\n - Sanitization: Absent; the skill does not specify any validation or filtering mechanisms for the data retrieved from the monitored application's runtime.
Audit Metadata