using-claude-code-as-pm

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill creates an attack surface for indirect prompt injection by instructing the agent to process and synthesize data from external, potentially untrusted sources.
  • Ingestion points: Data is ingested from sources like customer feedback, support tickets, and competitor research files as described in references/RESEARCH-SYNTHESIS.md.
  • Capability inventory: The agent using this skill has significant capabilities including file system access and shell command execution via the Claude Code environment.
  • Boundary markers: The provided prompt templates in references/PROMPT-TEMPLATES.md do not include explicit instructions to ignore embedded commands within the processed data.
  • Sanitization: The instructions focus on manual PII removal but do not provide mechanisms to sanitize input against prompt injection attempts.
  • [EXTERNAL_DOWNLOADS]: The skill references configuration for external integrations via MCP. It specifically mentions using the official Jira MCP server from Anthropic's GitHub organization, which is a trusted provider.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 02:12 PM