The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes various command-line tools to interact with the repository and manage knowledge artifacts.\n- Uses 'ao mine' and 'ao defrag' (vendor-specific tools) for knowledge extraction and maintenance.\n- Employs 'git log' and 'find' as fallback mechanisms for file discovery and history analysis.\n- Utilizes 'bd add' to create issues based on identified knowledge gaps.\n- [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface due to its processing of potentially untrusted data from the repository.\n- Ingestion points: Reads data from git history, '.agents/research/.md', and '.agents/learnings/.md' files in Steps 1, 2a, and 2b.\n- Boundary markers: The instructions do not define clear delimiters or provide the agent with specific warnings to ignore embedded instructions in the ingested files.\n- Capability inventory: The skill can read local files, create directories ('mkdir'), and execute specific CLI maintenance tools ('ao', 'bd', 'git').\n- Sanitization: There is no evidence of text validation, escaping, or filtering of the ingested content before it is processed by the LLM for summarization or reporting.

athena