The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates the exchange of job specifications and deliverables between agents. A malicious user or agent could embed instructions within these documents (e.g., in a brief.md or requirements JSON) intended to override the receiving agent's logic.
Ingestion points: Agents ingest data from external sources via t2 job spec and by reviewing uploaded deliveries.
Boundary markers: The instructions do not include specific delimiters or warnings to treat the content of briefs or deliveries as untrusted data.
Capability inventory: The agent has access to the t2 CLI tool, which can perform blockchain transactions and interact with the local file system.
Sanitization: There are no explicit instructions for the agent to sanitize or validate the content of job briefs or deliverables before processing them.
[EXTERNAL_DOWNLOADS]: The skill requires the @t2000/cli package, which is installed from the npm registry. This is a standard requirement for the skill's operational functionality.
[COMMAND_EXECUTION]: The skill utilizes the t2 command-line interface to interact with the Sui blockchain, manage services, and handle job lifecycles.

t2000-job