huggingface-backup
Safely back up local folders to private Hugging Face Buckets using the installed hf CLI and the hf-buckets Pi extension. Use when the user asks to upload/archive/back up local data, training data, agent traces, model artifacts, checkpoints, corpora, or folders to Hugging Face storage, especially when privacy, buckets, quarantine, resume, monitoring, or deleting local copies is involved.
What this skill does
# Hugging Face Backup Use this skill for safe Hugging Face bucket backups. First load/use the official `hf-cli` skill when you need current command details. It is installed at: `~/.agents/skills/hf-cli/SKILL.md` ## Golden Rules - Prefer **Hugging Face Buckets** for raw folder backups unless the user explicitly asks for a model/dataset/space repo. - Create buckets as **private by default**. - Never upload obvious secrets if avoidable. - Never permanently delete local data after upload. Move verified uploads to quarantine instead. - Verify remote bucket metadata and nonzero file count before quarantine. - Use conservative concurrency. HF/Xet can rate-limit hard; prefer one or two active syncs for multi-GB folders. - If an upload appears stuck with zero CPU, old logs, `429 Too Many Requests`, or `CLOSE_WAIT`, stop/restart with lower concurrency or serial sync. ## Safe Excludes Always include these excludes for bucket syncs unless the user explicitly overrides: ```bash --exclude '**/.git/**' --exclude '**/.DS_Store' --exclude '**/__pycache__/**' --exclude '**/.pytest_cache/**' --exclude '**/.venv/**' --exclude '**/venv/**' --exclude '**/.env' --exclude '**/.env.*' --exclude '**/*cookie*' --exclude '**/*Cookie*' --exclude '**/*token*' --exclude '**/*Token*' --exclude '**/*secret*' --exclude '**/*Secret*' --exclude '**/*.ovpn' --exclude '**/auth_session.json' ``` ## Pi Extension The global Pi extension is installed at: `~/.pi/agent/extensions/hf-buckets/index.ts` After `/reload` or a new Pi session, it provides: ### Slash commands ```bash /hf-backup <path> --bucket <bucket-name> [--namespace <user-or-org>] [--dry-run] /hf-status [job-id] /hf-quarantine <job-id> ``` ### Tools available to the model - `hf_safe_bucket_backup` — start a detached private bucket sync with secret-safe excludes. - `hf_backup_status` — inspect one or all backup jobs. - `hf_quarantine_uploaded` — move a completed verified source folder to quarantine. Job state lives under: `~/.pi/hf-backups/jobs/` Default quarantine root: `~/.trash/hf-uploaded-training-folders/` ## Manual Fallback Workflow If the extension is not loaded, use the official `hf` CLI directly: ```bash hf buckets create <namespace>/<bucket> --private --exist-ok hf buckets sync <local-path> hf://buckets/<namespace>/<bucket> \ --exclude '**/.git/**' \ --exclude '**/.DS_Store' \ --exclude '**/__pycache__/**' \ --exclude '**/.pytest_cache/**' \ --exclude '**/.venv/**' \ --exclude '**/venv/**' \ --exclude '**/.env' \ --exclude '**/.env.*' \ --exclude '**/*cookie*' \ --exclude '**/*Cookie*' \ --exclude '**/*token*' \ --exclude '**/*Token*' \ --exclude '**/*secret*' \ --exclude '**/*Secret*' \ --exclude '**/*.ovpn' \ --exclude '**/auth_session.json' hf buckets info <namespace>/<bucket> --format json hf buckets list <namespace>/<bucket> -R --quiet | wc -l ``` Only after the bucket is private and has nonzero files should local data be moved to quarantine: ```bash mkdir -p ~/.trash/hf-uploaded-training-folders/YYYY-MM-DD mv <local-path> ~/.trash/hf-uploaded-training-folders/YYYY-MM-DD/<bucket-name> ``` ## Monitoring Guidance For long-running uploads: - Start detached jobs or use the extension. - Monitor process status and per-job logs. - Use a subagent monitor for multi-bucket or multi-hour upload sessions. - Mark as stalled if log mtime is older than ten minutes and the process has zero CPU, especially if Xet logs show `429 Too Many Requests`. - Resume safely: `hf buckets sync` is resumable/deduplicating enough for retry-style workflows. ## Common Bucket Naming Use lowercase, hyphenated bucket names: - `project-backup` - `training-data-archive` - `model-artifacts-YYYY-MM-DD` - `research-corpus-YYYY-MM-DD` - `agent-traces-YYYY-MM-DD` If the user says "repo" but also says "bucket", prefer bucket. Ask only if ambiguous and destructive or public/private semantics matter.
Related in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.