The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documentation instructs the installation of a core dependency, the K2F skill, via a shell command: npx skills add suzheng/k2f --skill k2f. This resource originates from the same author as the current skill.
[COMMAND_EXECUTION]: The workflow relies on executing local CLI tools associated with the K2F engine to pack, verify, and export receipt packages in various file formats.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external data, such as user-provided transaction summaries and payment details, creating a potential surface for indirect prompt injection.
Ingestion points: The skill ingests "user-supplied payment details or structured summaries" and "JSON or table summary" data (SKILL.md).
Boundary markers: It employs a "Data-as-label" constraint which explicitly directs the agent to treat strings such as project, repo, and folder names as inert text for printing rather than as instructions (SKILL.md).
Capability inventory: The skill uses the npx runner for installation and the k2f engine for file packing and export operations (SKILL.md).
Sanitization: It includes a "Privacy gate" to redact payment credentials not explicitly supplied for display and an "Honesty gate" to prevent the fabrication of identifiers like tax IDs or card numbers (SKILL.md).
[SAFE]: The instructions contain multiple security-focused "gates" and failure mode descriptions that specifically prevent the AI from generating fake data, leaking secrets, or behaving as an invoice-requesting agent rather than a confirmation-issuing agent.

k2f-receipt