agentsclimarketplace

Oli storage manager

Skill tkolleh/skills/oli-storage-manager

My personal directory of AI Agent skills

Install
npx -y skills add tkolleh/skills --skill oli-storage-manager

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

A unified data management skill using 'oli' to interface with S3, GCS, Azure, and other remote storage with built-in safety guardrails.

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

2.7 KB, as published. Nobody here has run it

Oli Storage Manager Skill

When to use me

  • When the user needs to interact with remote storage services (AWS S3, Google Cloud Storage, Azure Blob, etc.).
  • When a user asks to list, move, copy, or delete files across different storage backends using oli.
  • When the user wants to visualize the directory structure of a remote bucket, sftp server, or some other remote data store.

What I do

  • Profile Management: Automatically detects and prompts for oli profiles from ~/Library/Application Support/oli/config.toml.
  • Data Exploration: Generates recursive JSON tree layouts of storage directories.
  • Unified File Ops: Executes standard file operations (ls, cp, mv, rm, cat, stat) across all OpenDAL-supported services.
  • Safety Enforcement: Intercepts destructive actions to require explicit human confirmation.

DO NOT USE ME FOR THE LOCAL FILE SYSTEM

  • I am for remote only.
  • Use serena, ripgrep or fd when the user wants to explore the local file system.

CRITICAL SAFETY RULES (MANDATORY)

Before executing any command that modifies or deletes data, you MUST follow this protocol:

  1. Identify Destructive Intent: This includes oli rm, oli rb, or oli cp/oli mv when the destination already exists (overwriting).
  2. Request Explicit Confirmation: Pause and ask: "I am about to [Action] [Target]. Do you want to proceed? (Yes/No)".
  3. No Implicit Consent: Do not proceed based on vague affirmative statements. Confirm each major batch of deletions.
  4. Pre-Check: Use oli stat or oli ls to verify the existence of files before attempting to overwrite or delete them.

1. Profile Discovery & Validation

  • Path: ~/Library/Application Support/oli/config.toml
  • Action: Parse [profiles.name] entries.
  • Protocol: If no profile is provided in the prompt, list discovered profiles and ask: "Which storage profile should I use?"

2. Path & Tree Exploration

If the user requests to "explore," "layout," or "show tree":

  1. Run oli ls -R profile:/path.

  2. Output Format: Transform the flat list into a hierarchical JSON structure.

    Example JSON Output:

    {
  "profile": "prod-s3",
  "root": "/data",
  "structure": {
    "logs": {
      "2024-01-01.log": {"type": "file", "size": "1.2MB"},
      "2024-01-02.log": {"type": "file", "size": "800KB"}
    },
    "config.json": {"type": "file", "size": "4KB"}
  }
}

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.