The Agent Skills Directory

[SAFE]: The skill's primary function is to organize development workflows into actionable, bite-sized tasks using standard tools like git and pytest.
[SAFE]: No obfuscation, data exfiltration patterns, or unauthorized credential access were detected in the skill instructions or associated prompts.
[PROMPT_INJECTION]: The skill demonstrates a surface for indirect prompt injection by ingesting untrusted external specifications to generate implementation plans containing code and shell commands.
Ingestion points: Processes external 'specs' or 'requirements' provided in the user context to define the scope and content of implementation plans.
Boundary markers: Absent. There are no explicit instructions to the agent to treat input data as untrusted or to ignore embedded instructions within the spec.
Capability inventory: Writes generated plans to the local filesystem (docs/superpowers/plans/), produces git commands (git add, git commit), and generates test execution commands (pytest). It also dispatches subagents via superpowers:subagent-driven-development and superpowers:executing-plans.
Sanitization: Absent. The skill does not define methods for escaping or validating the content of the input specification before it is interpolated into the plan structure.

writing-plans