Provide persistent network filesystem storage with natural-language search for AI agents.
npx skills add https://github.com/mem9-ai/drive9 --skill drive9
Please help me install this Agent Skill. Skill: drive9 Source: https://github.com/mem9-ai/drive9/tree/main/site Command: npx skills add https://github.com/mem9-ai/drive9 --skill drive9
drive9 lets you persist files across agent handoffs and session resets, so you can store outputs once and retrieve them later using natural-language search instead of brittle filenames or forgotten context.
:/data/file.txt).grep for meaning-based retrieval and find for exact structural filters like tags, dates, and sizes.Use case: store a set of project notes, datasets, and generated reports during an agent run, then later ask for “the pricing strategy draft” and retrieve the exact files even if you don’t remember the filename.
Ask your agent to install drive9, provision a workspace, and then search by meaning: run curl -fsSL https://drive9.ai/install.sh | sh && drive9 create && drive9 fs grep "pricing strategy" /.
High-intent search queries and answers about installing and using this skill.
Hybrid retrieval combines semantic search and keyword filters to locate files across projects. You run `grep` for meaning-based retrieval and `find` for exact structural filters like tags, dates, and sizes, allowing precise file discovery across directories.
Workspace provisioning creates isolated storage for AI agents to keep credentials and config locally. You run `drive9 create` after installation to establish a persistent network filesystem, enabling repeat access to stored outputs across agent handoffs and session resets.
Zero-copy linking performs metadata-only copies and renames for fast, space-efficient file duplication. Instead of moving actual file contents, it manipulates metadata within the network filesystem to rapidly duplicate or restructure files without consuming additional storage space.
Storing agent outputs across multiple projects requires a persistent network filesystem with hybrid semantic search. You upload files once using filesystem-style commands, then retrieve them later across projects and directories using natural-language queries instead of relying on exact filenames.
A root-name entry is required for discovery in the network filesystem. It uses an SKILL.md entry to establish the root directory structure, enabling the semantic search and indexing mechanisms to properly locate and retrieve files across isolated workspaces.
vessuxCreate an isolated Git worktree workspace for feature development.
joshuacookSync local and remote workspaces with Motoko framework changes.
ccfCodify Nx workspace patterns for monorepo setup, boundary enforcement, and caching.
moongci38-ossPerform semantic and hybrid vector plus BM25 search on document collections.