The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read and interpret test reports located in the .axis/reports/ directory. These reports contain rationales and summaries generated by LLM judges during agent testing, which constitutes a surface for processing potentially untrusted or external content.
Ingestion points: The agent is directed to read .axis/reports/<reportId>/report.json and per-scenario detail JSON files.
Boundary markers: The data is processed as structured JSON, which helps separate data from instructions.
Capability inventory: The skill utilizes filesystem read access to retrieve report data.
Sanitization: The instructions emphasize reading values verbatim from the manifest to prevent hallucinations or misinterpretations.
[COMMAND_EXECUTION]: The skill provides the agent with specific CLI commands for initializing projects (axis init), running tests (axis run), and managing baselines (axis baseline). These are standard operations for a testing framework.
[EXTERNAL_DOWNLOADS]: The skill references the official documentation site (axis.run) and provides installation commands for Netlify-authored tools. These resources align with the vendor's own infrastructure and are used for their intended primary purpose.

using-axis