The Agent Skills Directory

[DATA_EXFILTRATION]: The skill directs the agent to access credentials from the local file path ~/.build-host/credentials.json for authentication with the build.host service.
The operating procedure explicitly instructs the agent to never print API keys, OAuth tokens, or secrets in chat logs or source code, mitigating the risk of exposure.
[COMMAND_EXECUTION]: The skill requires the agent to perform local environment checks and interact with the build.host API via HTTP POST and GET requests for project deployment and log retrieval.
These operations are scoped to the primary purpose of the skill and include instructions for manual verification of returned URLs.
[PROMPT_INJECTION]: The skill ingests untrusted data from git tags and project repositories, which presents a surface for indirect prompt injection.
Ingestion points: External data enters the context through git tags and local project directory contents as specified in SKILL.md.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands within the tag data are present.
Capability inventory: The skill possesses capabilities for local file system operations and network API calls to the build.host infrastructure.
Sanitization: No specific validation or sanitization logic for the content of git tags or repository files is defined in the instructions.

changelog