inbound-triage

Pass

Audited by Gen Agent Trust Hub on Aug 4, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions for the installation and configuration of the paperclipai CLI and the @bbengamin/paperclip-mcp-server via npm. These tools are provided by the skill author to enable interaction with the Paperclip management environment.
  • [COMMAND_EXECUTION]: The instructions direct the agent to use shell commands through the paperclipai CLI to query and update organizational records. This is a core functionality of the skill for task management.
  • [PROMPT_INJECTION]: The skill processes inbound content from issue descriptions and wiki pages which are untrusted external inputs. This creates an indirect prompt injection surface. The evidence chain includes: 1. Ingestion points: Issue fields (status, plan, description) and wiki content fetched via paperclip-wiki-fetch. 2. Boundary markers: The skill strictly requires human operator approval before any mutation ('Mutation Rule'). 3. Capability inventory: Commands include issue updates, status changes, and wiki management via paperclipai and MCP tools. 4. Sanitization: Relies on human-in-the-loop verification of all proposed changes.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 4, 2026, 08:26 PM
Security Audit — agent-trust-hub — inbound-triage