The Agent Skills Directory

[SAFE]: The skill consists primarily of documentation and architectural patterns for task management within the Claude Code environment. These patterns encourage structured workflows and do not introduce security vulnerabilities.
[SAFE]: The script scripts/task-tree-visualizer.py is a benign utility for generating ASCII representations of task dependency trees. It uses standard Python libraries and does not perform network operations, file system modifications, or process execution.
[INDIRECT_PROMPT_INJECTION]: The skill defines WebFetch and WebSearch in its allowed-tools, which represents an ingestion surface for untrusted external data. However, the skill lacks dangerous capabilities that would make this surface exploitable, and it does not contain instructions that bypass safety guardrails.
Ingestion points: WebFetch and WebSearch tools used to retrieve external content.
Boundary markers: None explicitly defined in the pattern instructions for the web tools.
Capability inventory: The skill allows file reading (Read, Glob, Grep) but does not grant file-writing or arbitrary command execution capabilities.
Sanitization: Standard model-level sanitization and tool-level restrictions apply to the ingested content.

task-dependency-patterns