The Agent Skills Directory

[SAFE]: The skill uses a comprehensive template to guide the agent's logic, prioritizing verification through investigation of local code and official documentation.
[SAFE]: It implements robust uncertainty tracking by mandating 'Unproven' and 'Accepted risk' labels for claims that lack definitive evidence.
[SAFE]: The skill is entirely instructional and does not include any executable code, scripts, or remote network calls.
[SAFE]: An indirect prompt injection surface is noted as the skill processes local workspace data (Ingestion points: code, logs, and issues referenced in SKILL.md; Boundary markers: absent; Capability inventory: plans for file and shell access; Sanitization: absent), but this is inherent to its primary purpose as a planning tool.
[SAFE]: No obfuscation, persistence mechanisms, or credential harvesting patterns were found in the instructions.

vibe-planning