cy-create-techspec

Pass

Audited by Gen Agent Trust Hub on May 11, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core function of ingesting and processing untrusted external data.
  • Ingestion points: The agent reads content from _prd.md files, existing Architecture Decision Records (ADRs) in the .compozy/tasks/ directory, and performs general codebase exploration to analyze architecture and dependencies.
  • Boundary markers: Absent. The instructions do not define specific delimiters or escaping mechanisms to separate the data found in files from the agent's internal instruction context.
  • Capability inventory: The skill is capable of directory creation, writing multiple ADR markdown files, and writing the final TechSpec file to the local file system.
  • Sanitization: Absent. There are no instructions to validate, sanitize, or filter the content retrieved from the PRDs or source code before it is interpolated into the design drafts.
  • Process Controls: The skill includes significant procedural mitigations, such as a mandatory '' requiring user approval of the final draft and an interactive review phase where the user must approve the approach before any files are saved.
Audit Metadata
Risk Level
SAFE
Analyzed
May 11, 2026, 11:08 AM
Security Audit — agent-trust-hub — cy-create-techspec