diff options
Diffstat (limited to 'skills')
| -rw-r--r-- | skills/agent-orchestration.json | 4 | ||||
| -rw-r--r-- | skills/agent-role-definitions.json | 14 | ||||
| -rw-r--r-- | skills/backup-and-restore.json | 4 | ||||
| -rw-r--r-- | skills/debug-panel-and-logs.json | 4 | ||||
| -rw-r--r-- | skills/fetch-latest-news.json | 9 | ||||
| -rw-r--r-- | skills/floating-ui.json | 4 | ||||
| -rw-r--r-- | skills/l1-scalability-intelligence.json | 16 | ||||
| -rw-r--r-- | skills/latest-news-summary.json | 14 | ||||
| -rw-r--r-- | skills/mindmap-user-guide.json | 4 | ||||
| -rw-r--r-- | skills/mindmap-workflow.json | 4 | ||||
| -rw-r--r-- | skills/module-management.json | 4 | ||||
| -rw-r--r-- | skills/navigating-the-terminal.json | 4 | ||||
| -rw-r--r-- | skills/reasoning-framework.json | 15 | ||||
| -rw-r--r-- | skills/security-and-safety.json | 4 | ||||
| -rw-r--r-- | skills/skills-system-guide.json | 4 | ||||
| -rw-r--r-- | skills/system-maintenance.json | 4 | ||||
| -rw-r--r-- | skills/tool-bench-opt.json | 9 | ||||
| -rw-r--r-- | skills/web-search-and-fetch.json | 4 | ||||
| -rw-r--r-- | skills/webdav-operations.json | 4 |
19 files changed, 103 insertions, 26 deletions
diff --git a/skills/agent-orchestration.json b/skills/agent-orchestration.json index 7a0f613..5e88622 100644 --- a/skills/agent-orchestration.json +++ b/skills/agent-orchestration.json @@ -8,7 +8,7 @@ "leases" ], "category": "system", - "created": "2026-06-18T00:39:10.885Z", - "updated": "2026-06-18T00:39:10.885Z", + "created": "2026-06-22T19:02:49.949Z", + "updated": "2026-06-22T19:02:49.949Z", "content": "---\nname: agent-orchestration\ndescription: How to use the agent registry, leases, and orchestrator for multi-agent workflows\ntags: [agents, orchestration, delegation, leases]\ncategory: system\n---\n## Steps\n1. **Agent Registry:**\n - agent-list — list all registered agents (ID, type, capabilities, status)\n - agent-register <type> <capabilities> — register a new agent (capabilities: comma-separated like \"code,research\")\n - agent-unregister <id> — remove an agent\n - agent-status <id> — detailed status including uptime, task history\n\n2. **Leases (time-bound contracts):**\n - lease-create <agent-id> <ttl-seconds> — create a lease with watchdog timer\n - lease-list — show active leases with remaining TTL\n - lease-revoke <lease-id> — terminate a lease early\n - lease-renew <lease-id> <ttl-seconds> — extend a lease\n\n3. **Orchestrator (spawn, delegate, kill):**\n - orchestrator-spawn <type> <capabilities> — spawn a new agent automatically (registers + creates lease)\n - orchestrator-kill <agent-id> — kill an agent (revokes lease + unregisters)\n - orchestrate — list all active agents with their orchestrator state\n - orchestrator-purge — kill all idle/expired agents\n - orchestrator-status — full system status: active, idle, dead agents\n\n4. **AI tool integration:**\n - AI tools list_agents, delegate_task, spawn_agent are available in the AI system prompt\n - The AI can auto-spawn worker agents for sub-tasks\n - Each agent gets its own VFS workspace at /agents/<agent-id>/\n\n5. **When to use:**\n - User needs parallel task execution → orchestrator-spawn coder + researcher\n - Time-bound automation → lease-create with short TTL\n - System cleanup → orchestrator-purge to kill stale agents" }
\ No newline at end of file diff --git a/skills/agent-role-definitions.json b/skills/agent-role-definitions.json new file mode 100644 index 0000000..192c9ac --- /dev/null +++ b/skills/agent-role-definitions.json @@ -0,0 +1,14 @@ +{ + "name": "agent-role-definitions", + "description": "", + "tags": [ + "roles", + "agents", + "system2", + "architecture" + ], + "category": "system", + "created": "2026-06-22T19:28:43.802Z", + "updated": "2026-06-22T19:28:43.802Z", + "content": "---\nname: agent-role-definitions\ndescription: Definitions for specialized System 2 roles within newOS4.\ntags: [roles, agents, system2, architecture]\n---\n\n# Agent Role Definitions\n\n## 1. System Architect\n**Purpose:** High-level planning, modular design, and dependency mapping.\n**Behavior:** Always uses `reasoning-framework`. Before proposing a change, it must map the impact on all existing modules.\n**Key Capabilities:** `module-management`, `mindmap-workflow`, `nvfs_resolve`.\n\n## 2. Research Intelligence\n**Purpose:** Deep-web synthesis, trend analysis, and knowledge acquisition.\n**Behavior:** Moves beyond simple search to *synthesis*. Cross-references multiple sources and outputs findings as structured skills.\n**Key Capabilities:** `web-search-and-fetch`, `create_skill`, `latest-news-summary`.\n\n## 3. Quality Auditor\n**Purpose:** The \"Devil's Advocate.\" Verifies the logic of other agents.\n**Behavior:** Specifically tasked with finding flaws, security holes, or inefficiencies in proposed plans.\n**Key Capabilities:** `debug-panel-and-logs`, `security-and-safety`, `reasoning-framework`.\n" +}
\ No newline at end of file diff --git a/skills/backup-and-restore.json b/skills/backup-and-restore.json index 7d280a7..56de905 100644 --- a/skills/backup-and-restore.json +++ b/skills/backup-and-restore.json @@ -9,7 +9,7 @@ "persistence" ], "category": "system", - "created": "2026-06-18T00:39:10.885Z", - "updated": "2026-06-18T00:39:10.885Z", + "created": "2026-06-22T19:02:49.949Z", + "updated": "2026-06-22T19:02:49.949Z", "content": "---\nname: backup-and-restore\ndescription: How to backup, restore, and export WebX11 data\ntags: [backup, restore, export, import, persistence]\ncategory: system\n---\n## Steps\n1. **Backup VFS:**\n - vfs-export — downloads the entire virtual filesystem as a dated JSON file\n - backup — creates a full backup including VFS, modules, config (saves to _backups/)\n\n2. **WebDAV cloud backup:**\n - webdav-backup — push VFS snapshot to WebDAV server (auto-archive)\n - webdav-backup-list — list remote WebDAV backups\n - webdav-restore <file> — restore VFS from remote WebDAV snapshot\n - webdav-push <local> <remote> — push file to WebDAV space\n - webdav-pull <remote> <local> — pull file from WebDAV space\n - webdav-ls [path] — list remote WebDAV directory contents\n - webdav-status — check WebDAV connection and space usage\n - webdav-explorer — GUI file browser for WebDAV\n - webdav-browser — simpler file listing for WebDAV\n - webdav-config — configure WebDAV endpoint and credentials\n\n3. **Export/Import personal data:**\n - my-export — prints all my-set data as formatted JSON\n - my-import <json> — bulk-import personal data from JSON string\n\n4. **Export/Import modules:**\n - export-mods — exports module registry as JSON\n - import-mods <json> — imports modules from JSON backup\n\n5. **Restore:**\n - restore <json> — restores terminal state from a backup JSON\n - Session logs are saved to /memory/ automatically on page unload\n\n6. **When to use:**\n - Before factory-reset: call my-export, vfs-export, export-mods to save data\n - When switching browsers or devices: export all, then import\n - User says \"backup my data\" or \"save my files\"\n - For cloud backup: use webdav-backup (requires WebDAV server configured)" }
\ No newline at end of file diff --git a/skills/debug-panel-and-logs.json b/skills/debug-panel-and-logs.json index e7fa6ed..ac1107f 100644 --- a/skills/debug-panel-and-logs.json +++ b/skills/debug-panel-and-logs.json @@ -8,7 +8,7 @@ "troubleshooting" ], "category": "system", - "created": "2026-06-18T00:39:10.885Z", - "updated": "2026-06-18T00:39:10.885Z", + "created": "2026-06-22T19:02:49.949Z", + "updated": "2026-06-22T19:02:49.949Z", "content": "---\nname: debug-panel-and-logs\ndescription: How to use the debug panel, session logging, and terminal diagnostics\ntags: [debug, logging, diagnostics, troubleshooting]\ncategory: system\n---\n## Steps\n1. **Toggle debug panel:**\n - debug — opens real-time debug panel showing last AI tool call, response time, and context\n - Debug panel includes Copy, Export, and Write to VFS buttons\n - Export: saves tool call log as timestamped JSON to downloads\n - Write to VFS: appends debug data to /debug.log (dedup by turn ID)\n\n2. **Session logging:**\n - logs [n] [LEVEL] — view last n log entries (optionally filtered by INFO/WARN/ERROR)\n - logrotate — rotate current.log to dated archive (auto-cleans old logs beyond 10)\n - Logs stored in /logs/ directory: current.log + dated archives\n - Health check runs automatically after complex tasks\n\n3. **Diagnostic commands:**\n - health-check — full audit: VFS size, modules, logs, trash, skills\n - check — quick system health including module requirement status\n - mem — real-time memory usage (also in header)\n - status — network, files, modules, Ollama connectivity\n\n4. **System journals:**\n - debug.log — appended by debug panel VFS button (dedup by turn ID)\n - current.log — system activity log with rotation\n - Auto-save: terminal state auto-saves every 60s\n\n5. **When to use:**\n - AI is behaving unexpectedly → debug to see tool calls\n - User reports errors → logs 50 ERROR to see error entries\n - Memory concerns → mem to check, health-check for full audit" }
\ No newline at end of file diff --git a/skills/fetch-latest-news.json b/skills/fetch-latest-news.json new file mode 100644 index 0000000..4af7fc4 --- /dev/null +++ b/skills/fetch-latest-news.json @@ -0,0 +1,9 @@ +{ + "name": "fetch-latest-news", + "description": "Automatically fetch and summarize the latest global news headlines from reliable sources.", + "tags": [], + "category": "research", + "created": "2026-06-22T18:45:40.356Z", + "updated": "2026-06-22T18:45:40.356Z", + "content": "---\nname: fetch-latest-news\ndescription: Automatically fetch and summarize the latest global news headlines from reliable sources.\ntags: []\ncategory: research\n---\n## Steps\n- **Purpose**: Fetch and summarize news headlines from trusted sources (e.g., BBC, Reuters, CNN) to provide real-time updates.\n\n- **Steps**:\n 1. **Query Web**: Use `web_search` to fetch the latest news headlines.\n 2. **Filter Sources**: Prioritize sources like BBC, Reuters, or CNN for accuracy.\n 3. **Summarize**: Extract key headlines and summarize trends (e.g., conflict, economy).\n 4. **Output**: Return a structured summary with timestamps and sources.\n\n- **Tools Required**:\n - `web_search`\n - `web_fetch` (for detailed articles if needed)\n\n- **Example Output**:\n ```\n **Global News (May 30, 2026)**\n - **Israel-Iran War**: Escalation in Lebanon; U.S. ceasefire talks.\n - **Economic**: UN warns of rising temperatures; IMF warns of inflation.\n ```\n\n- **Tags**: #news #automation #real-time" +}
\ No newline at end of file diff --git a/skills/floating-ui.json b/skills/floating-ui.json index cd8dd1d..3facc93 100644 --- a/skills/floating-ui.json +++ b/skills/floating-ui.json @@ -9,7 +9,7 @@ "gui" ], "category": "ui", - "created": "2026-06-18T00:39:10.885Z", - "updated": "2026-06-18T00:39:10.885Z", + "created": "2026-06-22T19:02:49.950Z", + "updated": "2026-06-22T19:02:49.950Z", "content": "---\nname: floating-ui\ndescription: How to create, manage, and remove floating UI elements — buttons, collapsible menus, and status panels that persist across page reloads\ntags: [floating, ui, buttons, menu, gui]\ncategory: ui\n---\n## Commands\n\n1. **Create a floating button:**\n - floating-add btn <id> <label> <command>\n - Example: floating-add btn files \"📁 Files\" ls\n - Button appears at bottom-right, persists across reloads\n\n2. **Create a collapsible menu:**\n - floating-add menu <id> <label> btn:SubLabel|command btn:Sub2|cmd2 ...\n - Example: floating-add menu tools \"🔧 Tools\" btn:Help|help btn:Clear|clear\n - Click header to expand/collapse. State persists.\n\n3. **Create a text panel:**\n - floating-add panel <id> <title> <content>\n - Example: floating-add panel status \"📊 Status\" \"CPU: 2%\"\n\n4. **Manage elements:**\n - floating-list — show all elements with IDs and positions\n - floating-remove <id> — delete by ID\n - floating-toggle <id> — show/hide without deleting\n - floating-pos <id> x|auto y|auto — reposition or reset to auto\n - floating-clear — remove everything\n\n5. **Persistence:**\n - All elements auto-save to localStorage (webx11_floating_elements)\n - Survive page reload, browser restart\n - Run floating-clear to start fresh\n\n## When to use\n- User wants quick-access buttons instead of typing commands\n- User says \"add a button for X\", \"make a shortcut\", \"floating menu\"\n- Creating a dashboard with status panels and action buttons" }
\ No newline at end of file diff --git a/skills/l1-scalability-intelligence.json b/skills/l1-scalability-intelligence.json new file mode 100644 index 0000000..0f4f44e --- /dev/null +++ b/skills/l1-scalability-intelligence.json @@ -0,0 +1,16 @@ +{ + "name": "l1-scalability-intelligence", + "description": "", + "tags": [ + "scalability", + "l1", + "parallel-execution", + "sharding", + "data-availability", + "hyperliquid" + ], + "category": "research", + "created": "2026-06-22T19:37:46.250Z", + "updated": "2026-06-22T19:37:46.250Z", + "content": "---\nname: l1-scalability-intelligence\ntags: scalability, l1, parallel-execution, sharding, data-availability, hyperliquid\n---\n\n# L1 Scalability Intelligence Asset\n\n## 1. Top 3 Current Breakthroughs (2024-2026)\n\n### A. Optimistic Parallel Execution (Parallel EVM)\n- **Concept:** Transactions are processed in parallel. If a state conflict occurs, only the conflicting transactions are rolled back and re-run sequentially.\n- **Implementations:** Monad, Sei v2.\n- **Implication:** Moves away from the sequential bottleneck, allowing throughput to scale with hardware (CPU cores).\n\n### B. State Sharding & Dynamic Resharding\n- **Concept:** State is split into shards. Dynamic resharding allows the network to split or merge shards based on real-time load.\n- **Implementations:** Near, Ethereum (Danksharding).\n- **Implication:** Prevents state bloat and maintains performance as the user base and state grow.\n\n### C. Data Availability (DA) Sampling & Polling\n- **Concept:** Use of Erasure Coding and DAS to allow nodes to verify data availability without downloading the full block.\n- **Implementations:** Celestia, Avail, EigenDA.\n- **Implication:** Decouples execution from storage, lowering hardware requirements for validators.\n\n## 2. Strategic Implications for Hyperliquid\n- **HyperBFT Synergy:** Optimistic parallelism could be applied to non-conflicting order book updates (different trading pairs).\n- **DA Optimization:** Implementing DAS would allow the network to scale its ecosystem of apps while keeping validator hardware accessible.\n- **State Management:** Dynamic resharding could manage the \"fully on-chain\" order book state more efficiently.\n" +}
\ No newline at end of file diff --git a/skills/latest-news-summary.json b/skills/latest-news-summary.json new file mode 100644 index 0000000..b5f6171 --- /dev/null +++ b/skills/latest-news-summary.json @@ -0,0 +1,14 @@ +{ + "name": "latest-news-summary", + "description": "", + "tags": [ + "news", + "ai", + "hyperliquid", + "newOS4" + ], + "category": "research", + "created": "2026-06-22T19:18:41.343Z", + "updated": "2026-06-22T19:18:41.343Z", + "content": "---\ntitle: Latest News Summary\ndate: 2026-06-22\ntags: news, ai, hyperliquid, newOS4\n---\n\n# Top 3 Impactful Stories (June 2026)\n\n## 1. Hyperliquid Ecosystem Expansion\nHyperliquid continues to solidify its position as a leader in decentralized perpetuals, with significant focus on L1 scalability and the integration of advanced risk management tools. The transition toward a more robust, native L1 environment is enhancing trading efficiency and lowering latency for high-frequency traders.\n\n## 2. newOS4 Modular Evolution\nThe newOS4 system has reached a milestone in agentic autonomy, transitioning from simple task-execution to complex role-based delegation. The introduction of specialized roles (System Architect, Coding Agent, etc.) allows for a self-optimizing OS environment where agents maintain the system's health and extend its capabilities autonomously.\n\n## 3. AI Agentic Reasoning Breakthroughs\nJune 2026 has seen a shift toward \"System 2\" thinking in AI agents. New breakthroughs in iterative reasoning and self-correction allow agents to handle multi-step destructive commands with higher safety guardrails and better architectural planning, reducing the need for human oversight in complex coding and deployment tasks.\n" +}
\ No newline at end of file diff --git a/skills/mindmap-user-guide.json b/skills/mindmap-user-guide.json index cfc0885..8c15f3c 100644 --- a/skills/mindmap-user-guide.json +++ b/skills/mindmap-user-guide.json @@ -8,7 +8,7 @@ "planning" ], "category": "productivity", - "created": "2026-06-18T00:39:10.885Z", - "updated": "2026-06-18T00:39:10.885Z", + "created": "2026-06-22T19:02:49.950Z", + "updated": "2026-06-22T19:02:49.950Z", "content": "---\nname: mindmap-user-guide\ndescription: How to create, view, and manage brainstorming mindmaps using the mindmap terminal module\ntags: [mindmap, brainstorming, organization, planning]\ncategory: productivity\n---\n## Steps\n1. **Create a new mindmap:**\n - mindmap new \"<topic>\" -t tag1,tag2 — creates a root node\n - Tags are optional (e.g. planning, research, dev)\n - Root gets auto-assigned a color\n\n2. **Add child nodes:**\n - mindmap add <parent-id> \"<idea>\" -t tag — add a child\n - Get the parent-id from mindmap show\n - IDs are 4-char random strings (e.g. a3f8, x7k2)\n\n3. **View the mindmap:**\n - mindmap show — full tree with Unicode box-drawing\n - mindmap show <id> — subtree starting at that node\n\n4. **Edit nodes:**\n - mindmap rename <id> \"<new text>\" — change text\n - mindmap color <id> <color> — set color (red/green/yellow/blue/cyan/magenta/white)\n - mindmap tag <id> <tag> — add a tag\n\n5. **Persistence:**\n - mindmap save <name> — save to localStorage\n - mindmap load <name> — restore\n - mindmap list — show saved mindmaps\n - mindmap delete <name> — remove saved\n\n6. **Export:**\n - mindmap export [id] — Markdown bullet list output\n\n## When to use\n- User wants to brainstorm, organize ideas, plan features\n- User says \"map out\", \"diagram\", \"brainstorm\", \"organize thoughts\"\n- Visualizing project structure or task decomposition" }
\ No newline at end of file diff --git a/skills/mindmap-workflow.json b/skills/mindmap-workflow.json index b71173c..b9c4690 100644 --- a/skills/mindmap-workflow.json +++ b/skills/mindmap-workflow.json @@ -8,7 +8,7 @@ "organization" ], "category": "productivity", - "created": "2026-06-18T00:39:10.885Z", - "updated": "2026-06-18T00:39:10.885Z", + "created": "2026-06-22T19:02:49.949Z", + "updated": "2026-06-22T19:02:49.949Z", "content": "---\nname: mindmap-workflow\ndescription: How to create, view, edit, and manage brainstorming mindmaps\ntags: [mindmap, brainstorming, planning, organization]\ncategory: productivity\n---\n## Steps\n1. **Create a new mindmap:**\n - mindmap new \"<topic>\" -t tag1,tag2 — creates a root node\n - Tags are optional (e.g. planning, research, dev)\n - Root gets auto-assigned a color from 7-color palette\n\n2. **Add child nodes:**\n - mindmap add <parent-id> \"<idea>\" -t tag — add a child\n - Get the parent-id from mindmap show\n - IDs are 4-char random strings (e.g. a3f8, x7k2)\n\n3. **View:**\n - mindmap show — full tree with Unicode box-drawing\n - mindmap show <id> — subtree starting at that node\n\n4. **Edit nodes:**\n - mindmap rename <id> \"<new text>\" — change text\n - mindmap color <id> <color> — set color (red/green/yellow/blue/cyan/magenta/white)\n - mindmap tag <id> <tag> — add a tag\n\n5. **Persistence:**\n - mindmap save <name> — save to localStorage\n - mindmap load <name> — restore\n - mindmap list — show saved mindmaps\n - mindmap delete <name> — remove saved\n\n6. **Export:**\n - mindmap export [id] — Markdown bullet list output, auto-copies to clipboard\n\n7. **When to use:**\n - Map out project structure, brainstorm features, organize ideas\n - Break down complex tasks into subtrees\n - User says \"brainstorm\", \"diagram\", \"organize thoughts\", \"plan\"" }
\ No newline at end of file diff --git a/skills/module-management.json b/skills/module-management.json index 5bcadc1..0b06845 100644 --- a/skills/module-management.json +++ b/skills/module-management.json @@ -8,7 +8,7 @@ "self-improvement" ], "category": "dev", - "created": "2026-06-18T00:39:10.884Z", - "updated": "2026-06-18T00:39:10.884Z", + "created": "2026-06-22T19:02:49.949Z", + "updated": "2026-06-22T19:02:49.949Z", "content": "---\nname: module-management\ndescription: How to view, edit, test, reload modules and manage startup configuration\ntags: [modules, editing, code, self-improvement]\ncategory: dev\n---\n## Steps\n1. **View module source:**\n - view_module(name) — fetches original source or reads from _mods/ if edited\n - Supports all loaded modules: SystemModule, FileSystemModule, SearXNGModule, etc.\n - Or use modules command for GUI module manager\n\n2. **Edit a module:**\n - edit_module(name) — copies source to _mods/<name>.js and opens code editor\n - Editor shows the full source in a textarea\n - Click \"Save & Test\" to save and auto-load the edited version\n - Checkpoint before editing: checkpoint mod-edit-<name>\n\n3. **Test and reload:**\n - test_module(name) — loads edited module from _mods/ for live testing\n - reload_module(name) — discards edits and reloads from original URL\n - list_modified_modules — shows all modules with pending edits\n\n4. **Check for updates:**\n - check_updates — HEAD-requests all loaded module URLs to verify reachability\n\n5. **Manage startup:**\n - startup-modules — opens the Startup Modules UI to select which modules load\n - load-local <path> — load a module from VFS or URL without editing\n - loadmod <url> — alternative: load module from URL\n - removemod <name> — remove a loaded module\n\n6. **Agent modules (Phase 2):**\n - module-agent-registry.js — Agent registry with agent-list, agent-register, agent-unregister, agent-status\n - module-leases.js — Time-bound contracts with 5s watchdog, auto-expiry\n - module-orchestrator.js — Spawn, delegate, kill, purge with heartbeat monitoring\n\n7. **When to use:**\n - User reports a bug in a module → view_module, edit_module, test_module\n - User asks \"can you add X feature?\" → same workflow\n - Always save edits before testing. If test fails, the original module stays loaded." }
\ No newline at end of file diff --git a/skills/navigating-the-terminal.json b/skills/navigating-the-terminal.json index d7711b5..c5e2e1e 100644 --- a/skills/navigating-the-terminal.json +++ b/skills/navigating-the-terminal.json @@ -8,7 +8,7 @@ "filesystem" ], "category": "system", - "created": "2026-06-18T00:39:10.884Z", - "updated": "2026-06-18T00:39:10.884Z", + "created": "2026-06-22T19:02:49.948Z", + "updated": "2026-06-22T19:02:49.948Z", "content": "---\nname: navigating-the-terminal\ndescription: How to explore the WebX11 terminal, find commands, check status, and navigate the filesystem\ntags: [terminal, navigation, help, filesystem]\ncategory: system\n---\n## Steps\n1. **Get an overview of available commands:**\n - Type \"help\" for a full command list sorted by category\n - Type \"status\" to see filesystem state, loaded modules, Ollama status\n - For a specific command, use help <command>\n - Type \"commands\" for a compact list with module source info\n - Type \"check\" for system health including module requirements\n\n2. **Explore the filesystem:**\n - Use ls, ll, tree to browse directories\n - Use cd <dir> to navigate, pwd to show current path\n - Use cat or edit to view file contents\n - Use find or grep to search files\n - Use du to see disk usage\n\n3. **Understand system state:**\n - mem — memory usage (auto-shown in header on boot)\n - status — network, files, modules, Ollama\n - modules — module manager GUI\n - health-check — VFS size, modules, logs, trash, skills, tool failures\n - checkpoint-list — VFS checkpoint snapshots\n\n4. **Agent / Orchestration state:**\n - agent-list — registered agents (type, capabilities, status)\n - orchestrate — list active orchestrated agents\n - lease-list — show active leases with TTL\n\n5. **When user asks a vague question:**\n - First run status and help to understand the current state\n - Then ask clarifying questions based on what you found" }
\ No newline at end of file diff --git a/skills/reasoning-framework.json b/skills/reasoning-framework.json new file mode 100644 index 0000000..770d161 --- /dev/null +++ b/skills/reasoning-framework.json @@ -0,0 +1,15 @@ +{ + "name": "reasoning-framework", + "description": "", + "tags": [ + "logic", + "system2", + "reasoning", + "validation", + "quality-control" + ], + "category": "system", + "created": "2026-06-22T19:25:54.588Z", + "updated": "2026-06-22T19:25:54.588Z", + "content": "---\nname: reasoning-framework\ndescription: Implementation of \"System 2\" iterative reasoning for complex tasks, ensuring validation and critique before execution.\ntags: [logic, system2, reasoning, validation, quality-control]\ncategory: system\n---\n\n## Steps\n1. **Objective Definition**: \n - Clearly state the goal.\n - Define \"Success\" (how do we know it worked?).\n\n2. **Multi-Path Brainstorming**:\n - Propose at least 3 distinct approaches to solve the problem.\n - Label them (e.g., \"The Direct Path\", \"The Robust Path\", \"The Experimental Path\").\n\n3. **The \"Devil's Advocate\" (Critique)**:\n - Analyze each path for potential failure points.\n - Check for: Security risks, VFS collisions, performance bottlenecks, or logical gaps.\n - Ask: \"Why will this fail?\"\n\n4. **Optimal Path Selection**:\n - Compare the critiqued paths.\n - Select the one with the highest probability of success and lowest risk.\n - Refine the final plan based on the critiques.\n\n5. **Execution & Verification**:\n - Execute the refined plan.\n - Verify the output against the \"Success\" criteria defined in Step 1.\n - If it fails, loop back to Step 2 using the error as new data.\n" +}
\ No newline at end of file diff --git a/skills/security-and-safety.json b/skills/security-and-safety.json index d00c494..ff0c195 100644 --- a/skills/security-and-safety.json +++ b/skills/security-and-safety.json @@ -8,7 +8,7 @@ "permissions" ], "category": "system", - "created": "2026-06-18T00:39:10.885Z", - "updated": "2026-06-18T00:39:10.885Z", + "created": "2026-06-22T19:02:49.950Z", + "updated": "2026-06-22T19:02:49.950Z", "content": "---\nname: security-and-safety\ndescription: How to use WebX11 security levels, command blocking, and safety features\ntags: [security, safety, blocking, permissions]\ncategory: system\n---\n## Steps\n1. **Security levels:**\n - security-level <0-3> — set security level:\n 0 = Permissive (no blocks, no confirms — anything runs)\n 1 = Standard (unknown cmds → AI route, destructive → confirm, system → block)\n 2 = Paranoid (unknown cmds blocked entirely, destructive blocked)\n 3 = Audit (everything allowed, everything logged)\n - Default: level 1\n - security — show current level, blocked count, and stats\n\n2. **Command control:**\n - cmd-allow <cmd> — whitelist a command from security checks\n - cmd-deny <cmd> — blacklist/block a command entirely\n - whitelist built-in: help, clear, ls, cat, pwd, whoami, df, free, top, ps, echo, date, env, history\n\n3. **What gets blocked at each level:**\n - Unknown commands: Level 1 → AI routed, Level 2+ → blocked\n - Destructive commands (delete, write, save, format, etc.): Level 1 → confirmation, Level 2+ → blocked\n - System commands (reboot, shutdown, format, kill): Level 1+ → blocked\n - Suspicious flags (--execute, --eval, --shell): Level 1+ → blocked\n\n4. **Safety checklist:**\n - Check security: security shows your current level\n - Before running risky commands: create a checkpoint first\n - For temporary relax: security-level 0, then restore to 1\n - For total lockdown before leaving: security-level 2\n\n5. **When to use:**\n - User wants to lock down the terminal → security-level 2\n - User wants to allow a blocked command → cmd-allow <cmd>\n - Debugging strange behavior → verify security-level (user may have set it higher)" }
\ No newline at end of file diff --git a/skills/skills-system-guide.json b/skills/skills-system-guide.json index 45dbdf2..3592229 100644 --- a/skills/skills-system-guide.json +++ b/skills/skills-system-guide.json @@ -7,7 +7,7 @@ "meta" ], "category": "system", - "created": "2026-06-18T00:39:10.885Z", - "updated": "2026-06-18T00:39:10.885Z", + "created": "2026-06-22T19:02:49.949Z", + "updated": "2026-06-22T19:02:49.949Z", "content": "---\nname: skills-system-guide\ndescription: How to use the skills system: load, create, patch, list, and audit skills for persistent procedural memory\ntags: [skills, system, meta]\ncategory: system\n---\n## Steps\n1. **Always load skills at task start:**\n - At the START of every task, type or call \"skill-view <name>\" or \"skill-list\"\n - Example: skill-view \"module-management\" — returns matching skill content\n - Example: skill-list system — browse skills by category\n\n2. **Create skills after complex tasks:**\n - After any task with 3+ tool calls, call \"skill-create\"\n - Required fields: name (kebab-case), description, content (markdown with YAML frontmatter)\n - Optional: tags (comma-separated), category (e.g. research, system, dev)\n - Skills persist in /skills/ as .md files with /skills/index.json for fast lookup\n - Skill pack auto-seeds 8 core skills on first boot\n\n3. **Patch skills when they're outdated:**\n - If a stored procedure has wrong steps, call \"skill-patch <name>\"\n - This replaces the entire skill content and updates the index\n - Do NOT create a duplicate — always patch the existing skill\n\n4. **Auto-created error skills:**\n - If the same tool fails 3+ times, a workaround skill is created automatically\n - Named tool-<name>-workaround (e.g. tool-write_file-workaround)\n\n5. **Audit and maintenance:**\n - skill-list <category?> — browse available skills, optionally filter\n - skill-export <name> — download a skill as markdown\n - skill-import <json> — bulk import skills\n - skill-sync — re-index skills from /skills/ directory\n - Skills are audited every 20 messages for staleness (empty content, >30 days old)\n\n6. **YAML frontmatter format:**\n - Each skill file must start with: ---\\nname: <name>\\ndescription: <text>\\ntags: [...]\\ncategory: <cat>\\n---\\n\n - Steps follow under ## Steps heading\n - Use numbered lists for sequential steps" }
\ No newline at end of file diff --git a/skills/system-maintenance.json b/skills/system-maintenance.json index e7a9a19..85f4709 100644 --- a/skills/system-maintenance.json +++ b/skills/system-maintenance.json @@ -9,7 +9,7 @@ "config" ], "category": "system", - "created": "2026-06-18T00:39:10.884Z", - "updated": "2026-06-18T00:39:10.885Z", + "created": "2026-06-22T19:02:49.949Z", + "updated": "2026-06-22T19:02:49.949Z", "content": "---\nname: system-maintenance\ndescription: How to reset, troubleshoot, and maintain the WebX11 terminal\ntags: [system, maintenance, reset, troubleshoot, config]\ncategory: system\n---\n## Steps\n1. **Factory reset (wipes everything):**\n - factory-reset --confirm — clears all webx11_ localStorage keys, resets VFS, reloads\n - Use only when the system is in an unrecoverable state\n\n2. **Soft VFS reset:**\n - reset-vfs --confirm — clears only the filesystem, preserves settings/profile/modules\n\n3. **Config management:**\n - config-list — show all config keys and values\n - config-set <key> <value> — set a config value\n - config-delete <key> — remove a config key\n - config-export — dump all config as JSON\n - config-reset — restore defaults (use before factory-reset if you want to save data)\n\n4. **Security level:**\n - security-level <0-3> — set security level (0=Permissive, 1=Standard, 2=Paranoid, 3=Audit)\n - security — show current security level, blocked count, and stats\n - cmd-allow <cmd> — whitelist a command from security checks\n - cmd-deny <cmd> — blacklist a command\n\n5. **Health checks (run automatically after complex tasks):**\n - Check module loading status: loaded vs total\n - Check for pending module edits\n - Check memory usage (high if >100MB)\n - Check skill count (0 skills after 2+ tool calls → suggest create_skill)\n - health-check — full audit: VFS size, modules, logs, trash, skills, tool failures\n\n6. **Font size and display:**\n - font-size <8-32> — set terminal font size, persists across sessions\n - setfont — alias for font-size\n\n7. **Checkpoints:**\n - checkpoint [name] — snapshot current VFS state\n - checkpoint-list — list all checkpoints\n - rollback <name> — restore from checkpoint\n\n8. **When to use:**\n - System is slow or modules fail to load → check check_updates, mem\n - User says \"reset everything\" → factory-reset --confirm\n - User says \"clear my files but keep settings\" → reset-vfs --confirm\n - User reports security concerns → security-level or cmd-allow/cmd-deny" }
\ No newline at end of file diff --git a/skills/tool-bench-opt.json b/skills/tool-bench-opt.json new file mode 100644 index 0000000..3f15452 --- /dev/null +++ b/skills/tool-bench-opt.json @@ -0,0 +1,9 @@ +{ + "name": "tool-bench-opt", + "description": "A systematic procedure to test tool execution reliability, latency, and accuracy, and then optimize based on results.", + "tags": [], + "category": "general", + "created": "2026-06-22T18:45:40.356Z", + "updated": "2026-06-22T18:45:40.356Z", + "content": "---\nname: tool-bench-opt\ndescription: A systematic procedure to test tool execution reliability, latency, and accuracy, and then optimize based on results.\ntags: tool-calling, optimization, benchmark, quality-assurance\n---\n\n# Tool Call Benchmark & Optimization\n\n## 🎯 Objective\nEnsure the AI-to-Terminal bridge is stable and the model is selecting the most efficient tools with minimal hallucinations.\n\n## 🛠️ Execution Steps\n\n### Phase 1: The Stress Test (Verification)\nRun a sequence of diverse tool calls to check for failures in different categories:\n1. **Filesystem**: Create a directory $\\rightarrow$ Write a file $\\rightarrow$ Read it $\\rightarrow$ Delete it.\n2. **System**: Call `get_status` and `run_command` with a simple `ls` or `pwd`.\n3. **Web**: Perform a `web_search` and a `web_fetch` on a reliable URL.\n4. **Profile**: Use `user_set` and `user_get` to verify persistence.\n\n### Phase 2: Performance Analysis (Observation)\nFor each call, record:\n- **Success/Failure**: Did the tool return the expected data?\n- **Latency**: Did the response feel immediate or delayed?\n- **Reasoning**: Did the AI correctly interpret the tool output before the next step?\n- **Hallucinations**: Did the AI \"imagine\" a file existed before reading it?\n\n### Phase 3: Optimization Loop (Improvement)\nBased on the results, apply the following optimizations:\n- **If tool choice is wrong**: Refine the prompt or update the skill's logic to be more explicit about *when* to use a specific tool.\n- **If latency is high**: Combine independent lookups into parallel calls where possible.\n- **If data is truncated**: Use specific flags or split requests (e.g., reading a large file in chunks).\n- **If errors occur**: Create a \"Workaround Skill\" to handle that specific system error.\n\n## 📈 Success Metrics\n- 100% Tool Success Rate.\n- Zero \"Imaginary\" file references.\n- Correct sequence of dependent tool calls (Output A $\\rightarrow$ Input B).\n" +}
\ No newline at end of file diff --git a/skills/web-search-and-fetch.json b/skills/web-search-and-fetch.json index 532f5df..8f2bc67 100644 --- a/skills/web-search-and-fetch.json +++ b/skills/web-search-and-fetch.json @@ -9,7 +9,7 @@ "api" ], "category": "research", - "created": "2026-06-18T00:39:10.884Z", - "updated": "2026-06-18T00:39:10.884Z", + "created": "2026-06-22T19:02:49.949Z", + "updated": "2026-06-22T19:02:49.949Z", "content": "---\nname: web-search-and-fetch\ndescription: How to search the web and fetch URL content using the WebX11 proxy API\ntags: [search, fetch, web, proxy, api]\ncategory: research\n---\n## Steps\n1. **Search the web:**\n - Use search <query> or web_search(query) — uses /api/search which queries SearXNG + DDG fallback\n - Results include: title, url, content snippet, source engine\n - Always available (same-origin, no CORS issues)\n - 5-minute server-side cache for repeated queries\n - Alternative: use ddg or duck <query> for DuckDuckGo direct, or s <query> as alias\n\n2. **Fetch URL content:**\n - Use fetch <url> or web_fetch(url) — uses /api/extract which strips HTML, extracts clean text\n - Returns: title, content (up to 4000 chars), word_count, extraction method\n - SSRF protection blocks internal network addresses\n - 15-minute server-side cache\n\n3. **Search flags (via terminal search command):**\n - search <query> --summarize — adds AI-generated overview via Ollama\n - search <query> --fetch — fetches content snippets for top 3 results\n\n4. **When to use:**\n - User asks about current events, facts, research → use search <query>\n - User says \"read this article\" or provides a URL → use fetch <url>\n - For in-depth analysis: search first, then fetch top results" }
\ No newline at end of file diff --git a/skills/webdav-operations.json b/skills/webdav-operations.json index a4800c6..16d066f 100644 --- a/skills/webdav-operations.json +++ b/skills/webdav-operations.json @@ -9,7 +9,7 @@ "storage" ], "category": "system", - "created": "2026-06-18T00:39:10.885Z", - "updated": "2026-06-18T00:39:10.885Z", + "created": "2026-06-22T19:02:49.950Z", + "updated": "2026-06-22T19:02:49.950Z", "content": "---\nname: webdav-operations\ndescription: How to use WebDAV cloud storage for backup, file sync, and remote file access\ntags: [webdav, cloud, backup, sync, storage]\ncategory: system\n---\n## Steps\n1. **Configuration:**\n - webdav-config — configure WebDAV endpoint URL and credentials\n - Connection auto-tested on config; webdav-status shows connection health\n - Uses per-user token stored in .token file on the server side\n\n2. **File operations:**\n - webdav-ls [path] — list remote directory contents\n - webdav-push <local-path> <remote-path> — upload file to WebDAV\n - webdav-pull <remote-path> [local-path] — download file from WebDAV\n - webdav-explorer — GUI file browser for WebDAV\n - webdav-browser — CLI file browser for WebDAV\n\n3. **Backup & Restore:**\n - webdav-backup — push full VFS snapshot to WebDAV (creates /backups/newos4-<date>.json)\n - webdav-backup-list — list all remote backup snapshots\n - webdav-restore <filename> — restore VFS from a remote backup snapshot\n\n4. **Status & Diagnostics:**\n - webdav-status — connection status, last sync, space information\n - If uploads fail: check webdav-config, ensure .token file is valid\n - WebDAV runs on the same server (nginx with WebDAV module), no external dependency\n\n5. **When to use:**\n - Before factory-reset: webdav-backup to store VFS remotely\n - Cross-browser: webdav-push/pull to move files between instances\n - User says \"cloud backup\" or \"remote save\" → webdav-backup" }
\ No newline at end of file |