notion-spec-to-implementation

Pass

Audited by Gen Agent Trust Hub on May 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted data from external specification documents to drive automated task creation.
  • Ingestion points: Data enters the agent's context through the Notion:notion-fetch tool, which retrieves the full content of specification pages found during the Notion:notion-search phase, as documented in SKILL.md and reference/spec-parsing.md.
  • Boundary markers: There are no explicit delimiters or specific instructions to the agent to treat the fetched content as data only; the workflow assumes the content consists of valid requirements, which could allow an attacker to embed malicious instructions within a Notion page.
  • Capability inventory: The skill has broad write access to the user's Notion environment, including the ability to create new pages (Notion:notion-create-pages) and modify existing ones (Notion:notion-update-page).
  • Sanitization: The skill lacks any filtering, escaping, or validation mechanisms for the content retrieved from Notion before it is used to generate implementation plans and individual tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
May 17, 2026, 03:02 PM
Security Audit — agent-trust-hub — notion-spec-to-implementation