slack
Pass
Audited by Gen Agent Trust Hub on May 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
agent-browsertool within a Bash environment to perform complex browser automation tasks, including session connection, navigation, and interactive element manipulation. - [DATA_EXFILTRATION]: The skill is designed to capture and save Slack workspace data, including messages, user information, and visual snapshots (screenshots), to local files for further processing and reporting.
- [PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection because it processes untrusted data from an external source (Slack).
- Ingestion points: Slack message content, channel names, and user data are ingested into the agent context via
agent-browser snapshotandagent-browser get textcommands. - Boundary markers: Absent. The instructions do not provide delimiters or specific guidance for the agent to distinguish between valid data and potential malicious instructions embedded within Slack messages.
- Capability inventory: The agent has access to
Bashcommands and specialized browser tools (click,fill,press) that can modify the state of the Slack workspace or perform actions on behalf of the user. - Sanitization: Absent. There is no evidence of sanitization or validation of the extracted Slack content before it is processed by the agent.
Audit Metadata