skills/marcioaltoe/skills/write-tasks/Gen Agent Trust Hub

write-tasks

Pass

Audited by Gen Agent Trust Hub on Aug 26, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external specification files to generate task-related instructions, which constitutes an ingestion surface for potentially malicious data.
  • Ingestion points: The skill reads _prd.md and _techspec.md files from repository paths defined by user arguments (SKILL.md).
  • Boundary markers: The process uses structured Markdown templates and YAML frontmatter to delineate task metadata from instructional content (references/task-template.md).
  • Capability inventory: The skill generates task_NN.md files that include executable shell commands for environment and task verification (SKILL.md).
  • Sanitization: The skill enforces a "Project Constraint preflight" that requires explicit maintainer authorization recorded in specifications before any repository-tooling configuration can be modified. It also defines five specific escalation triggers to stop autonomous work when architecture, authority, or blast radius limits are exceeded (SKILL.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 26, 2026, 05:22 PM
Security Audit — agent-trust-hub — write-tasks