The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses curl to interact with Azure DevOps REST APIs. These commands are localized to data retrieval from a well-known service and are necessary for the skill's primary function.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data from Azure DevOps (test titles, error messages, and stack traces) which could theoretically be manipulated to include hidden instructions.
Ingestion points: Data is fetched via API calls to dev.azure.com and vstmr.dev.azure.com (SKILL.md).
Boundary markers: The skill does not explicitly define delimiters for external data, though it limits stack trace output to the first 20 lines.
Capability inventory: The skill is restricted to read-only network operations via curl and does not possess file-writing or code execution capabilities.
Sanitization: No explicit sanitization or escaping of the ingested API content is defined.
[CREDENTIALS_UNSAFE]: The skill manages authentication via the AZURE_DEVOPS_PAT environment variable. It includes explicit safety instructions for the agent to never print the token, write it to files, or ask the user to provide it directly in the chat, following secret management best practices.

pipelines-test-results