Add zo-goal-loop skill (jaknyfe)

Community/huggingface-hub/SKILL.md

@@ -0,0 +1,68 @@
+---
+name: huggingface-hub
+description: Access Hugging Face Hub models, datasets, and spaces via the huggingface_hub Python library. Use when you need to list, search, download, or upload HF assets.
+compatibility: Created for Zo Computer
+metadata:
+  author: jaknyfe.zo.computer
+---
+
+# Hugging Face Hub Skill
+
+Interact with Hugging Face Hub — browse models/datasets/spaces, download files, upload artifacts, and more.
+
+## Prerequisites
+
+Install the library:
+```bash
+pip install huggingface_hub
+```
+
+Save your HF token to [Settings > Advanced](/?t=settings&s=advanced) as `HF_TOKEN` (get one at https://huggingface.co/settings/tokens).
+
+## Usage
+
+```bash
+python3 /home/workspace/Skills/huggingface-hub/scripts/hf_hub.py <command> [options]
+```
+
+### Commands
+
+| Command | Description |
+|---------|-------------|
+| `list-models` | List models. Options: `--search`, `--sort`, `--direction`, `--limit` |
+| `list-datasets` | List datasets. Options: `--search`, `--sort`, `--direction`, `--limit` |
+| `list-spaces` | List spaces. Options: `--search`, `--sort`, `--direction`, `--limit` |
+| `model-info` | Get model info. Options: `--model` |
+| `dataset-info` | Get dataset info. Options: `--dataset` |
+| `download-file` | Download a file from a repo. Options: `--repo-id`, `--filename`, `--revision` |
+| `upload-file` | Upload a file. Options: `--local-path`, `--repo-id`, `--repo-type`, `--path-in-repo` |
+| `whoami` | Show authenticated user info |
+
+### Examples
+
+```bash
+# Search for text-to-image models
+python3 .../hf_hub.py list-models --search "text-to-image" --limit 10
+
+# Download a model file
+python3 .../hf_hub.py download-file --repo-id "stabilityai/stable-diffusion-xl-base-1.0" --filename "pytorch_model.bin"
+
+# Upload a file
+python3 .../hf_hub.py upload-file --local-path ./model.pt --repo-id "your-username/my-model" --repo-type "model"
+
+# List top-rated datasets
+python3 .../hf_hub.py list-datasets --sort "likes" --direction desc --limit 5
+```
+
+## API Reference
+
+This skill wraps `huggingface_hub`. Key functions:
+- `list_models`, `list_datasets`, `list_spaces` — browse hub
+- `huggingface_hub` — authentication and file ops
+- `InferenceClient` — run inference on hosted models/spaces
+
+## Notes
+
+- Downloads cache to `~/.cache/huggingface/`
+- `repo-type` values: `model`, `dataset`, `space`
+- Default revision is `main`

Community/huggingface-hub/scripts/hf_hub.py

@@ -0,0 +1,168 @@
+#!/usr/bin/env /usr/local/bin/python3
+"""Hugging Face Hub CLI tool. Usage: python3 hf_hub.py <command> [options]"""
+
+import argparse
+import os
+import sys
+from pathlib import Path
+
+try:
+    from huggingface_hub import (
+        HfApi, hf_hub_download,
+        list_models, list_datasets, list_spaces,
+        model_info, dataset_info,
+    )
+except ImportError:
+    sys.stderr.write("ERROR: huggingface_hub not installed. Run: pip install huggingface_hub\n")
+    sys.exit(1)
+
+
+def get_token() -> str:
+    token = os.environ.get("HF_TOKEN", "")
+    if not token:
+        try:
+            with open("/etc/secrets/HF_TOKEN", "r") as f:
+                token = f.read().strip()
+        except Exception:
+            pass
+    return token
+
+
+def get_api() -> HfApi:
+    token = get_token()
+    return HfApi(token=token if token else None)
+
+
+def cmd_whoami(args):
+    api = get_api()
+    info = api.whoami()
+    print(f"Username: {info['name']}")
+    print(f"Full name: {info.get('fullname', 'N/A')}")
+    print(f"Email: {info.get('email', 'N/A')}")
+    print(f"Organizations: {', '.join(info.get('orgs', []))}")
+
+
+def cmd_list_models(args):
+    kwargs = {"sort": args.sort, "direction": args.direction, "limit": args.limit}
+    if args.search:
+        kwargs["search"] = args.search
+    kwargs = {k: v for k, v in kwargs.items() if v is not None}
+    for m in list_models(**kwargs):
+        print(f"{m.id}  likes={m.likes}")
+
+
+def cmd_list_datasets(args):
+    kwargs = {"sort": args.sort, "direction": args.direction, "limit": args.limit}
+    if args.search:
+        kwargs["search"] = args.search
+    kwargs = {k: v for k, v in kwargs.items() if v is not None}
+    for d in list_datasets(**kwargs):
+        print(f"{d.id}  likes={d.likes}")
+
+
+def cmd_list_spaces(args):
+    kwargs = {"sort": args.sort, "direction": args.direction, "limit": args.limit}
+    if args.search:
+        kwargs["search"] = args.search
+    kwargs = {k: v for k, v in kwargs.items() if v is not None}
+    for s in list_spaces(**kwargs):
+        print(f"{s.id}  likes={s.likes}")
+
+
+def cmd_model_info(args):
+    info = model_info(args.model)
+    print(f"Model ID: {info.id}")
+    print(f"Downloads: {getattr(info, 'downloads', 'N/A')}")
+    print(f"Likes: {info.likes}")
+    print(f"Tags: {', '.join(info.tags)}")
+    print(f"Pipeline tag: {getattr(info, 'pipeline_tag', 'N/A')}")
+    print(f"Created at: {info.created_at}")
+    print(f"Last modified: {info.last_modified}")
+
+
+def cmd_dataset_info(args):
+    info = dataset_info(args.dataset)
+    print(f"Dataset ID: {info.id}")
+    print(f"Downloads: {getattr(info, 'downloads', 'N/A')}")
+    print(f"Likes: {info.likes}")
+    print(f"Tags: {', '.join(info.tags)}")
+    print(f"Created at: {info.created_at}")
+    print(f"Last modified: {info.last_modified}")
+
+
+def cmd_download_file(args):
+    path = hf_hub_download(
+        repo_id=args.repo_id,
+        filename=args.filename,
+        revision=args.revision or "main",
+        token=get_token() or None,
+    )
+    print(path)
+
+
+def cmd_upload_file(args):
+    api = get_api()
+    api.upload_file(
+        path_or_fileobj=args.local_path,
+        path_in_repo=args.path_in_repo or Path(args.local_path).name,
+        repo_id=args.repo_id,
+        repo_type=args.repo_type or "model",
+        token=get_token() or None,
+    )
+    print(f"Uploaded {args.local_path} to {args.repo_id}/{args.path_in_repo}")
+
+
+def build_parser():
+    parser = argparse.ArgumentParser(description="Hugging Face Hub CLI")
+    sub = parser.add_subparsers(dest="command")
+    sub.add_parser("whoami", help="Show authenticated user info")
+    p = sub.add_parser("list-models", help="List models")
+    p.add_argument("--search")
+    p.add_argument("--sort", default=None)
+    p.add_argument("--direction", default=None)
+    p.add_argument("--limit", type=int, default=20)
+    p = sub.add_parser("list-datasets", help="List datasets")
+    p.add_argument("--search")
+    p.add_argument("--sort", default=None)
+    p.add_argument("--direction", default=None)
+    p.add_argument("--limit", type=int, default=20)
+    p = sub.add_parser("list-spaces", help="List spaces")
+    p.add_argument("--search")
+    p.add_argument("--sort", default=None)
+    p.add_argument("--direction", default=None)
+    p.add_argument("--limit", type=int, default=20)
+    p = sub.add_parser("model-info", help="Get model info")
+    p.add_argument("--model", required=True)
+    p = sub.add_parser("dataset-info", help="Get dataset info")
+    p.add_argument("--dataset", required=True)
+    p = sub.add_parser("download-file", help="Download a file from a repo")
+    p.add_argument("--repo-id", required=True)
+    p.add_argument("--filename", required=True)
+    p.add_argument("--revision", default=None)
+    p = sub.add_parser("upload-file", help="Upload a file")
+    p.add_argument("--local-path", required=True)
+    p.add_argument("--repo-id", required=True)
+    p.add_argument("--repo-type", default="model")
+    p.add_argument("--path-in-repo", default=None)
+    return parser
+
+
+COMMANDS = {
+    "whoami": cmd_whoami,
+    "list-models": cmd_list_models,
+    "list-datasets": cmd_list_datasets,
+    "list-spaces": cmd_list_spaces,
+    "model-info": cmd_model_info,
+    "dataset-info": cmd_dataset_info,
+    "download-file": cmd_download_file,
+    "upload-file": cmd_upload_file,
+}
+
+
+if __name__ == "__main__":
+    parser = build_parser()
+    args = parser.parse_args()
+    if args.command is None:
+        parser.print_help()
+        sys.exit(1)
+    COMMANDS[args.command](args)

Community/zo-goal-loop/DISPLAY.json

@@ -0,0 +1,11 @@
+{
+  "name": "zo-goal-loop",
+  "displayName": "Zo Goal Loop",
+  "description": "Recreates Claude's /goal and /loop for autonomous task execution",
+  "category": "Utility",
+  "version": "1.0.0",
+  "scripts": {
+    "goal": "bun run /home/workspace/Skills/zo-goal-loop/scripts/goal.ts",
+    "loop": "bun run /home/workspace/Skills/zo-goal-loop/scripts/loop.ts"
+  }
+}

Community/zo-goal-loop/SKILL.md

@@ -0,0 +1,59 @@
+---
+name: zo-goal-loop
+description: Recreates Claude Code's /goal and /loop commands — define a completion condition and let Zo work autonomously until it's achieved. Supports iterative task execution with evaluator model checking for completion.
+metadata:
+  author: jaknyfe.zo.computer
+  category: Utility
+  display-name: Zo Goal Loop
+compatibility: Created for Zo Computer. Requires ZO_CLIENT_IDENTITY_TOKEN environment variable.
+---
+
+# Zo Goal Loop
+
+Run Zo autonomously until a goal is achieved.
+
+## Usage
+
+### /goal — Work until condition met
+
+```bash
+bun run /home/workspace/Skills/zo-goal-loop/scripts/goal.ts "completion condition here"
+```
+
+Examples:
+- `"All tests pass and lint is clean"`
+- `"Build completes successfully"`
+- `"Find 3 working solutions to the problem"`
+
+Optional args:
+- `--max-turns=<n>` — Stop after N turns (default: unlimited)
+- `--save-to=<file.md>` — Log progress to file
+
+### /loop — Run prompt repeatedly
+
+```bash
+bun run /home/workspace/Skills/zo-goal-loop/scripts/loop.ts "prompt" [options]
+```
+
+Options:
+- `--interval=<ms>` — Wait between iterations (default: 5000)
+- `--max-iterations=<n>` — Stop after N loops (default: unlimited)
+- `--until="<condition>"` — Stop when condition is met (file check, output match, etc.)
+
+## How it works
+
+1. `/goal` calls the Zo `/zo/ask` API with the task
+2. Each turn runs the task model
+3. After each turn, an evaluator model (Haiku) checks if the condition is met
+4. If not met, generates next action and continues
+5. If met, reports completion and any artifacts
+
+## Verification commands
+
+- Check progress: `cat /tmp/zo-goal-progress.log`
+- View current goal: `cat /tmp/current-goal.txt`
+- Force stop: `rm /tmp/current-goal.txt`
+
+## Requirements
+
+- `ZO_CLIENT_IDENTITY_TOKEN` must be available in environment