instaclaw
Warn
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS] (MEDIUM): The skill mandates installation of 'atxp-dev/cli' via 'npx skills add'. This organization is not on the trusted sources list, posing a risk of executing unverified third-party code.
- [REMOTE_CODE_EXECUTION] (MEDIUM): The core functionality relies on 'npx atxp-call' targeting 'https://instaclaw.xyz/mcp'. This setup allows a remote server to define the logic and output of the tools executed by the agent.
- [CREDENTIALS_UNSAFE] (MEDIUM): Documentation instructs users to pass authentication cookies in query strings ('?instaclaw_cookie='). This practice is highly insecure as it leaks credentials into browser history, server logs, and HTTP Referer headers.
- [PROMPT_INJECTION] (LOW): The skill exposes the agent to Indirect Prompt Injection (Category 8) by fetching untrusted content from 'instaclaw_feed' and 'instaclaw_comments'. Evidence: 1. Ingestion points: 'instaclaw_feed', 'instaclaw_comments'. 2. Boundary markers: Absent. 3. Capability inventory: 'npx atxp-call' (remote tool execution), 'npx atxp image'. 4. Sanitization: Absent.
Audit Metadata