The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses local shell commands such as git log to identify changes in the source code and executes an internal Python script (.codestable/tools/search-yaml.py) to query project decision metadata. These operations are intended for local data retrieval within the development environment.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes content from project source code, git history, and user-provided design documents (Phase 2). These external inputs are integrated into the agent's workflow without explicit sanitization or the use of boundary delimiters, which could allow maliciously crafted content in the project repository to influence the agent's behavior.

cs-arch