The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill recommends the installation of well-known Python packages including chromadb, sentence-transformers, and watchdog. These are standard, reputable libraries used for the skill's stated purpose of local vector storage and file monitoring.
[EXTERNAL_DOWNLOADS]: The included Python scripts utilize the sentence-transformers library, which downloads pre-trained embedding models (e.g., all-MiniLM-L6-v2) from Hugging Face, a well-known service in the machine learning community.
[PROMPT_INJECTION]: The skill is designed to ingest data from local Markdown files (wiki/**/*.md) and project memory paths (~/.claude/projects/*/memory/) into the agent's context for retrieval. This creates a surface for potential indirect prompt injection if the source files contain instructions that the agent might inadvertently follow when they are retrieved.
Ingestion points: Local markdown files located in the wiki/ directory and Claude's project memory directories.
Boundary markers: No specific delimiters or "ignore instructions" markers are implemented in the provided code snippets to isolate retrieved data.
Capability inventory: The skill possesses capabilities for file system read/write operations and local vector database management.
Sanitization: The skill does not perform sanitization or validation of the text content within the markdown files before it is processed for embeddings or retrieval.

knowledge-ops