The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from project repositories, which provides a surface for indirect prompt injection attacks where malicious content in a repository could attempt to override agent behavior.
Ingestion points: The skill uses trsd library read, trsd library search, trsd library query, and trsd library context to fetch data from repository files into the agent context (evidence in references/knowledge.md).
Capability inventory: The skill has the ability to write to workspaces, submit changes for review, and decide on publications using trsd library workspace write, trsd library workspace submit, and trsd library reviews decide (evidence in references/knowledge.md).
Boundary markers: The instructions do not define boundary markers or explicit warnings to the agent to ignore instructions embedded within the library content.
Sanitization: No evidence of input sanitization or schema validation for the data retrieved via the CLI tool is provided.
[COMMAND_EXECUTION]: The skill relies on the execution of the trsd CLI tool to perform its operations. These commands are executed locally to interact with the TreeSeed platform (evidence in references/knowledge.md).

treeseed