The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local commands such as openspec, git, and mkdir to manage the development workflow and tutorial steps. These are necessary for the skill's primary function.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) as it ingests untrusted data from the codebase when scanning for tasks. \n
Ingestion points: Codebase files scanned for strings like TODO, FIXME, and HACK in SKILL.md. \n
Boundary markers: No specific boundary markers or 'ignore' instructions are present for the ingested codebase content. \n
Capability inventory: Use of openspec CLI, file modifications, and git command execution. \n
Sanitization: No explicit sanitization or filtering of the scanned codebase content is mentioned.

openspec-onboard