setup-pstack

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from multiple sources including external model slugs detected from the environment, contents of existing configuration files (~/.claude/pstack-models.md, ~/.codex/config.toml), and direct user input.
  • Ingestion points: Reads from ~/.claude/pstack-models.md, ~/.codex/config.toml, and accepts user-pasted model slugs.
  • Boundary markers: None specified in the instructions for the generated markdown files.
  • Capability inventory: Writes to the file system (~/.claude/pstack-models.md, ~/.claude/CLAUDE.md) and interacts with subagents via Agent calls using these model slugs.
  • Sanitization: The instructions require validation against a detected set of slugs but do not specify sanitization for the content written to the markdown configuration files, which are later included as context.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 11:13 PM
Security Audit — agent-trust-hub — setup-pstack