The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the coval CLI to manage resources on the Coval platform, such as agents, personas, and test runs. These operations are essential for the skill's documented purpose and are executed within the user's authenticated environment.
[EXTERNAL_DOWNLOADS]: The skill references official installation instructions for the vendor's CLI and makes API requests to the vendor's domain (api.coval.dev). These are verified vendor resources.
[CREDENTIALS_UNSAFE]: Authentication for the API is handled through the use of an environment variable (COVAL_API_KEY), which is a recommended and safe method for managing secrets.
[PROMPT_INJECTION]: The skill processes data from the Coval API to build a report, presenting a surface for indirect prompt injection.
Ingestion points: JSON output from coval CLI commands processed in SKILL.md.
Boundary markers: Not present.
Capability inventory: coval CLI and curl commands in SKILL.md.
Sanitization: Uses jq to parse JSON and extract expected fields.

run-accent-testing