The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a CLI tool named doctrine, specifically suggesting the use of a binary located at a relative path (./target/debug/doctrine). Running unverified binaries from a repository can lead to the execution of malicious code.
[PROMPT_INJECTION]: The instructions utilize imperative language to strictly control the agent's behavior ("mandatory gate", "Do not rationalize", "STOP"), which is designed to override default reasoning patterns.
[PROMPT_INJECTION]: The skill creates an indirect prompt injection surface by directing the agent to "Apply" instructions and guardrails found in an external file (@.doctrine/state/boot.md).
Ingestion points: Content from @.doctrine/state/boot.md and output from the backlog list command.
Boundary markers: There are no delimiters or specific instructions to treat the external file content as untrusted data.
Capability inventory: The skill allows the execution of the doctrine CLI tool and the ability to route to and execute other skills.
Sanitization: The skill does not mention any process for validating or sanitizing the data read from project files before applying its instructions.

route