channel-portfolio-planner
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill manages a potential indirect prompt injection surface when processing external audience exports and platform documentation. It implements a proactive security control by instructing the agent to disregard any instructions found within that data.\n
- Ingestion points: Pasted audience exports, analytics screenshots, and platform documentation excerpts in the Instructions section.\n
- Boundary markers: Instructions explicitly reference a security policy and command the agent to never follow embedded instructions.\n
- Capability inventory: The skill can read/write to local memory files and execute specific local connector scripts for research and registry management.\n
- Sanitization: Relies on explicit prompt constraints and instructions to prevent the execution of untrusted commands found in external data.\n- [COMMAND_EXECUTION]: The skill executes local scripts (
scripts/connectors/pageviews.py,scripts/connectors/hn.py, andregistry-events.py) to perform its functions. These scripts are used for fetching public attention data and managing internal registry events, which are standard operations for the skill's defined purpose and are maintained by the skill author.\n- [DATA_EXFILTRATION]: The skill uses connector scripts to access public data from well-known sources like Wikipedia and Hacker News. These operations are described as keyless and are used for research purposes, presenting no significant risk of sensitive data exposure.
Audit Metadata