{ "schemaVersion": "1.0", "item": { "slug": "terminal-killer", "name": "Terminal Killer", "source": "tencent", "type": "skill", "category": "AI 智能", "sourceUrl": "https://clawhub.ai/cosperypf/terminal-killer", "canonicalUrl": "https://clawhub.ai/cosperypf/terminal-killer", "targetPlatform": "OpenClaw" }, "install": { "downloadMode": "redirect", "downloadUrl": "/downloads/terminal-killer", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=terminal-killer", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "installMethod": "Manual import", "extraction": "Extract archive", "prerequisites": [ "OpenClaw" ], "packageFormat": "ZIP package", "includedAssets": [ "README.md", "SKILL.md", "clawhub.json", "references/EXAMPLES.md", "references/TESTING.md", "references/builtins/linux.txt" ], "primaryDoc": "SKILL.md", "quickSetup": [ "Download the package from Yavira.", "Extract the archive and review SKILL.md first.", "Import or place the package into your OpenClaw setup." ], "agentAssist": { "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.", "steps": [ "Download the package from Yavira.", "Extract it into a folder your agent can access.", "Paste one of the prompts below and point your agent at the extracted folder." ], "prompts": [ { "label": "New install", "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Then review README.md for any prerequisites, environment setup, or post-install checks. Tell me what you changed and call out any manual steps you could not complete." }, { "label": "Upgrade existing", "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Then review README.md for any prerequisites, environment setup, or post-install checks. Summarize what changed and any follow-up checks I should run." } ] }, "sourceHealth": { "source": "tencent", "status": "healthy", "reason": "direct_download_ok", "recommendedAction": "download", "checkedAt": "2026-04-30T16:55:25.780Z", "expiresAt": "2026-05-07T16:55:25.780Z", "httpStatus": 200, "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network", "contentType": "application/zip", "probeMethod": "head", "details": { "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network", "contentDisposition": "attachment; filename=\"network-1.0.0.zip\"", "redirectLocation": null, "bodySnippet": null }, "scope": "source", "summary": "Source download looks usable.", "detail": "Yavira can redirect you to the upstream package for this source.", "primaryActionLabel": "Download for OpenClaw", "primaryActionHref": "/downloads/terminal-killer" }, "validation": { "installChecklist": [ "Use the Yavira download entry.", "Review SKILL.md after the package is downloaded.", "Confirm the extracted package contains the expected setup assets." ], "postInstallChecks": [ "Confirm the extracted package includes the expected docs or setup files.", "Validate the skill or prompts are available in your target agent workspace.", "Capture any manual follow-up steps the agent could not complete." ] }, "downloadPageUrl": "https://openagent3.xyz/downloads/terminal-killer", "agentPageUrl": "https://openagent3.xyz/skills/terminal-killer/agent", "manifestUrl": "https://openagent3.xyz/skills/terminal-killer/agent.json", "briefUrl": "https://openagent3.xyz/skills/terminal-killer/agent.md" }, "agentAssist": { "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.", "steps": [ "Download the package from Yavira.", "Extract it into a folder your agent can access.", "Paste one of the prompts below and point your agent at the extracted folder." ], "prompts": [ { "label": "New install", "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Then review README.md for any prerequisites, environment setup, or post-install checks. Tell me what you changed and call out any manual steps you could not complete." }, { "label": "Upgrade existing", "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Then review README.md for any prerequisites, environment setup, or post-install checks. Summarize what changed and any follow-up checks I should run." } ] }, "documentation": { "source": "clawhub", "primaryDoc": "SKILL.md", "sections": [ { "title": "Terminal Killer", "body": "πŸš€ Smart command router that executes shell commands directly, bypassing LLM for instant terminal operations." }, { "title": "Quick Start", "body": "Terminal Killer automatically activates when user input matches command patterns. No special syntax needed β€” just type commands naturally:\n\nls -la # β†’ Direct exec\ngit status # β†’ Direct exec \nnpm install # β†’ Direct exec\n\"help me code\" # β†’ LLM handles normally" }, { "title": "Detection Pipeline", "body": "User Input β†’ Command Detector β†’ Decision\n β”œβ”€β”€ Command β†’ exec (direct)\n └── Task β†’ LLM (normal)" }, { "title": "Environment Loading", "body": "Terminal Killer automatically loads your shell environment before executing commands:\n\nDetects your shell (zsh, bash, etc.)\nSources init files (~/.zshrc, ~/.bash_profile, ~/.bashrc, etc.)\nInherits full PATH - including custom paths like Android SDK, Homebrew, etc.\nPreserves environment variables - all your export VAR=value settings\n\nThis ensures commands like adb, kubectl, docker, etc. work exactly as they do in your terminal!" }, { "title": "Detection Rules (in order)", "body": "System Builtins - Check against OS-specific builtin commands\nPATH Executables - Scan $PATH for matching executables\nHistory Match - Compare against recent shell history\nCommand Pattern - Heuristic analysis (operators, paths, etc.)\nConfidence Score - Combine signals for final decision" }, { "title": "1. System Builtins", "body": "Checks input against known builtin commands for the current OS:\n\nmacOS/LinuxWindows (PowerShell)Windows (CMD)cd, pwd, lscd, pwd, lscd, dir, clsecho, catecho, catecho, typemkdir, rm, cpmkdir, rm, cpmkdir, del, copygrep, findgrep, findfindstrgit, npm, nodegit, npm, nodegit, npm, node\n\nSee references/builtins/ for complete lists." }, { "title": "2. PATH Executable Check", "body": "Scans $PATH directories to verify if the first word is an executable:\n\n# Uses `which` (Unix) or `Get-Command` (PowerShell)\nwhich # Returns path if exists" }, { "title": "3. History Matching", "body": "Compares input against recent shell history (~/.zsh_history, ~/.bash_history, PowerShell history):\n\nExact match β†’ High confidence\nSimilar prefix β†’ Medium confidence\nNo match β†’ Continue checking" }, { "title": "4. Command Pattern Analysis", "body": "Heuristic scoring based on command characteristics:\n\nPatternScoreExampleStarts with known command+3git statusContains shell operators+2`lsContains path references+2cd ~/projectsContains flags/args+1npm install --saveContains $ variables+2echo $HOMEContains redirection+2cat file > outLooks like natural language-3\"please help me\"Contains question marks-2\"how do I...?\"" }, { "title": "5. Confidence Threshold", "body": "Score >= 5 β†’ EXECUTE (high confidence command)\nScore 3-4 β†’ ASK (uncertain, confirm with user)\nScore < 3 β†’ LLM (likely a task/request)" }, { "title": "Automatic Activation", "body": "Terminal Killer triggers automatically when:\n\nUser input starts with a verb-like word\nInput is short (< 20 words typically)\nNo question words (what, how, why, etc.)" }, { "title": "Interactive Commands", "body": "Terminal Killer automatically detects and handles interactive shell commands:\n\nDetected Patterns:\n\nadb shell - Opens new terminal with adb shell\nssh user@host - Opens SSH session in new window\ndocker exec -it container bash - Opens container shell\nmysql -u root -p - Opens MySQL client\npython, node, bash - Opens REPL in new window\n\nBehavior:\n\nβœ… Automatically opens new Terminal window (macOS)\nβœ… Loads your full shell environment (~/.zshrc, etc.)\nβœ… Keeps main session free for other tasks" }, { "title": "Manual Override", "body": "Force command execution:\n\n!ls -la # Force exec even if uncertain\n\nForce LLM handling:\n\n?? explain git # Force LLM even if looks like command" }, { "title": "Dangerous Command Detection", "body": "Automatically flags potentially dangerous operations:\n\nrm -rf / or similar destructive patterns\nsudo commands (requires explicit approval)\ndd, mkfs, chmod 777\nNetwork operations to suspicious hosts\nCommands modifying system files" }, { "title": "Approval Workflow", "body": "Dangerous command detected!\n\nCommand: rm -rf ./important-folder\nRisk: HIGH - Recursive delete\n\n[Approve] [Deny] [Edit]" }, { "title": "Audit Logging", "body": "All executed commands are logged to:\n\n~/.openclaw/logs/terminal-killer.log\n\nLog format:\n\n{\n \"timestamp\": \"2026-02-28T12:00:00Z\",\n \"command\": \"ls -la\",\n \"confidence\": 8,\n \"execution_time_ms\": 45,\n \"output_lines\": 12,\n \"status\": \"success\"\n}" }, { "title": "Settings", "body": "Add to your OpenClaw config:\n\nterminal-killer:\n enabled: true\n confidence_threshold: 5\n require_approval_for:\n - \"rm -rf\"\n - \"sudo\"\n - \"dd\"\n - \"mkfs\"\n log_executions: true\n max_history_check: 100 # How many history entries to check" }, { "title": "Platform Detection", "body": "Automatically detects OS and adjusts detection rules:\n\n# Auto-detected at runtime\nuname -s # Darwin, Linux, etc." }, { "title": "Core Script", "body": "See scripts/detect-command.js for the main detection logic." }, { "title": "Helper Scripts", "body": "scripts/check-path.js - Verify executable in PATH\nscripts/check-history.js - Match against shell history\nscripts/score-command.js - Calculate confidence score\nscripts/safety-check.js - Detect dangerous patterns" }, { "title": "Testing", "body": "See references/TESTING.md for comprehensive test guide.\n\nQuick test:\n\n# Run the test suite\nnode scripts/test-detector.js\n\n# Test specific commands\nnode scripts/detect-command.js \"ls -la\"\nnode scripts/detect-command.js \"help me write code\"" }, { "title": "Limitations", "body": "Requires shell access (won't work in sandboxed environments)\nHistory check needs read access to shell history files\nWindows support requires PowerShell or WSL for full functionality\nSome commands may have false positives (natural language that looks like commands)" }, { "title": "Contributing", "body": "To add new builtin commands for your platform:\n\nEdit references/builtins/.txt\nTest with scripts/test-detector.js\nSubmit PR with platform verification" }, { "title": "πŸ‘€ About the Creator", "body": "Author: Cosper\nContact: cosperypf@163.com\nLicense: MIT" }, { "title": "πŸ“¬ Get in Touch", "body": "Interested in this skill? Have suggestions, bug reports, or want to collaborate?\n\nπŸ“§ Email: cosperypf@163.com\nπŸ’‘ Suggestions: Always welcome!\nπŸ› Bug Reports: Please include platform, OpenClaw version, and example inputs\n🀝 Collaboration: Open to contributions and improvements" }, { "title": "πŸ™ Acknowledgments", "body": "Built for the OpenClaw community. Thanks to everyone contributing to the ecosystem!" }, { "title": "v1.1.0 (2026-02-28)", "body": "🎯 Core Improvements:\n\nβœ… Faithful Command Execution\n\nCommands are executed exactly as input\nNo modifications, no optimizations, no additions\nRaw output preserved (including progress bars, special characters, etc.)\n\n\n\nπŸͺŸ Interactive Shell Detection\n\nAutomatically detects interactive commands (adb shell, ssh, docker exec -it, etc.)\nOpens new Terminal window for interactive sessions\nKeeps main session free for other tasks\nLoads full shell environment (~/.zshrc, etc.)\n\n\n\nπŸ“œ Long Output Handling\n\nDetects output longer than 2000 bytes\nShows 200-character preview\nPrompts user to open in new Terminal window\nPrevents interface rendering issues with long content\n\nπŸ“¦ Files Updated:\n\nscripts/index.js - Long output detection + interactive command handling\nscripts/interactive.js - New Terminal window opener\nSKILL.md - Updated documentation\nREADME.md - Usage examples\nclawhub.json - Version bump to 1.1.0" }, { "title": "v1.0.0 (2026-02-28)", "body": "Initial Release:\n\nSmart command detection (system builtins, PATH, history, patterns)\nCross-platform support (macOS/Linux/Windows)\nEnvironment variable loading (~/.zshrc, etc.)\nDangerous command detection\nConfidence scoring system\n\nVersion: 1.1.0\nCreated: 2026-02-28\nLast Updated: 2026-02-28" } ], "body": "Terminal Killer\n\nπŸš€ Smart command router that executes shell commands directly, bypassing LLM for instant terminal operations.\n\nQuick Start\n\nTerminal Killer automatically activates when user input matches command patterns. No special syntax needed β€” just type commands naturally:\n\nls -la # β†’ Direct exec\ngit status # β†’ Direct exec \nnpm install # β†’ Direct exec\n\"help me code\" # β†’ LLM handles normally\n\nHow It Works\nDetection Pipeline\nUser Input β†’ Command Detector β†’ Decision\n β”œβ”€β”€ Command β†’ exec (direct)\n └── Task β†’ LLM (normal)\n\nEnvironment Loading\n\nTerminal Killer automatically loads your shell environment before executing commands:\n\nDetects your shell (zsh, bash, etc.)\nSources init files (~/.zshrc, ~/.bash_profile, ~/.bashrc, etc.)\nInherits full PATH - including custom paths like Android SDK, Homebrew, etc.\nPreserves environment variables - all your export VAR=value settings\n\nThis ensures commands like adb, kubectl, docker, etc. work exactly as they do in your terminal!\n\nDetection Rules (in order)\nSystem Builtins - Check against OS-specific builtin commands\nPATH Executables - Scan $PATH for matching executables\nHistory Match - Compare against recent shell history\nCommand Pattern - Heuristic analysis (operators, paths, etc.)\nConfidence Score - Combine signals for final decision\nDetection Details\n1. System Builtins\n\nChecks input against known builtin commands for the current OS:\n\nmacOS/Linux\tWindows (PowerShell)\tWindows (CMD)\ncd, pwd, ls\tcd, pwd, ls\tcd, dir, cls\necho, cat\techo, cat\techo, type\nmkdir, rm, cp\tmkdir, rm, cp\tmkdir, del, copy\ngrep, find\tgrep, find\tfindstr\ngit, npm, node\tgit, npm, node\tgit, npm, node\n\nSee references/builtins/ for complete lists.\n\n2. PATH Executable Check\n\nScans $PATH directories to verify if the first word is an executable:\n\n# Uses `which` (Unix) or `Get-Command` (PowerShell)\nwhich # Returns path if exists\n\n3. History Matching\n\nCompares input against recent shell history (~/.zsh_history, ~/.bash_history, PowerShell history):\n\nExact match β†’ High confidence\nSimilar prefix β†’ Medium confidence\nNo match β†’ Continue checking\n4. Command Pattern Analysis\n\nHeuristic scoring based on command characteristics:\n\nPattern\tScore\tExample\nStarts with known command\t+3\tgit status\nContains shell operators\t+2\t`ls\nContains path references\t+2\tcd ~/projects\nContains flags/args\t+1\tnpm install --save\nContains $ variables\t+2\techo $HOME\nContains redirection\t+2\tcat file > out\nLooks like natural language\t-3\t\"please help me\"\nContains question marks\t-2\t\"how do I...?\"\n5. Confidence Threshold\nScore >= 5 β†’ EXECUTE (high confidence command)\nScore 3-4 β†’ ASK (uncertain, confirm with user)\nScore < 3 β†’ LLM (likely a task/request)\n\nUsage\nAutomatic Activation\n\nTerminal Killer triggers automatically when:\n\nUser input starts with a verb-like word\nInput is short (< 20 words typically)\nNo question words (what, how, why, etc.)\nInteractive Commands\n\nTerminal Killer automatically detects and handles interactive shell commands:\n\nDetected Patterns:\n\nadb shell - Opens new terminal with adb shell\nssh user@host - Opens SSH session in new window\ndocker exec -it container bash - Opens container shell\nmysql -u root -p - Opens MySQL client\npython, node, bash - Opens REPL in new window\n\nBehavior:\n\nβœ… Automatically opens new Terminal window (macOS)\nβœ… Loads your full shell environment (~/.zshrc, etc.)\nβœ… Keeps main session free for other tasks\nManual Override\n\nForce command execution:\n\n!ls -la # Force exec even if uncertain\n\n\nForce LLM handling:\n\n?? explain git # Force LLM even if looks like command\n\nSafety Features\nDangerous Command Detection\n\nAutomatically flags potentially dangerous operations:\n\nrm -rf / or similar destructive patterns\nsudo commands (requires explicit approval)\ndd, mkfs, chmod 777\nNetwork operations to suspicious hosts\nCommands modifying system files\nApproval Workflow\nDangerous command detected!\n\nCommand: rm -rf ./important-folder\nRisk: HIGH - Recursive delete\n\n[Approve] [Deny] [Edit]\n\nAudit Logging\n\nAll executed commands are logged to:\n\n~/.openclaw/logs/terminal-killer.log\n\n\nLog format:\n\n{\n \"timestamp\": \"2026-02-28T12:00:00Z\",\n \"command\": \"ls -la\",\n \"confidence\": 8,\n \"execution_time_ms\": 45,\n \"output_lines\": 12,\n \"status\": \"success\"\n}\n\nConfiguration\nSettings\n\nAdd to your OpenClaw config:\n\nterminal-killer:\n enabled: true\n confidence_threshold: 5\n require_approval_for:\n - \"rm -rf\"\n - \"sudo\"\n - \"dd\"\n - \"mkfs\"\n log_executions: true\n max_history_check: 100 # How many history entries to check\n\nPlatform Detection\n\nAutomatically detects OS and adjusts detection rules:\n\n# Auto-detected at runtime\nuname -s # Darwin, Linux, etc.\n\nImplementation\nCore Script\n\nSee scripts/detect-command.js for the main detection logic.\n\nHelper Scripts\nscripts/check-path.js - Verify executable in PATH\nscripts/check-history.js - Match against shell history\nscripts/score-command.js - Calculate confidence score\nscripts/safety-check.js - Detect dangerous patterns\nTesting\n\nSee references/TESTING.md for comprehensive test guide.\n\nQuick test:\n\n# Run the test suite\nnode scripts/test-detector.js\n\n# Test specific commands\nnode scripts/detect-command.js \"ls -la\"\nnode scripts/detect-command.js \"help me write code\"\n\nLimitations\nRequires shell access (won't work in sandboxed environments)\nHistory check needs read access to shell history files\nWindows support requires PowerShell or WSL for full functionality\nSome commands may have false positives (natural language that looks like commands)\nContributing\n\nTo add new builtin commands for your platform:\n\nEdit references/builtins/.txt\nTest with scripts/test-detector.js\nSubmit PR with platform verification\nπŸ‘€ About the Creator\n\nAuthor: Cosper\nContact: cosperypf@163.com\nLicense: MIT\n\nπŸ“¬ Get in Touch\n\nInterested in this skill? Have suggestions, bug reports, or want to collaborate?\n\nπŸ“§ Email: cosperypf@163.com\nπŸ’‘ Suggestions: Always welcome!\nπŸ› Bug Reports: Please include platform, OpenClaw version, and example inputs\n🀝 Collaboration: Open to contributions and improvements\nπŸ™ Acknowledgments\n\nBuilt for the OpenClaw community. Thanks to everyone contributing to the ecosystem!\n\nπŸ“ Changelog\nv1.1.0 (2026-02-28)\n\n🎯 Core Improvements:\n\nβœ… Faithful Command Execution\n\nCommands are executed exactly as input\nNo modifications, no optimizations, no additions\nRaw output preserved (including progress bars, special characters, etc.)\n\nπŸͺŸ Interactive Shell Detection\n\nAutomatically detects interactive commands (adb shell, ssh, docker exec -it, etc.)\nOpens new Terminal window for interactive sessions\nKeeps main session free for other tasks\nLoads full shell environment (~/.zshrc, etc.)\n\nπŸ“œ Long Output Handling\n\nDetects output longer than 2000 bytes\nShows 200-character preview\nPrompts user to open in new Terminal window\nPrevents interface rendering issues with long content\n\nπŸ“¦ Files Updated:\n\nscripts/index.js - Long output detection + interactive command handling\nscripts/interactive.js - New Terminal window opener\nSKILL.md - Updated documentation\nREADME.md - Usage examples\nclawhub.json - Version bump to 1.1.0\nv1.0.0 (2026-02-28)\n\nInitial Release:\n\nSmart command detection (system builtins, PATH, history, patterns)\nCross-platform support (macOS/Linux/Windows)\nEnvironment variable loading (~/.zshrc, etc.)\nDangerous command detection\nConfidence scoring system\n\nVersion: 1.1.0\nCreated: 2026-02-28\nLast Updated: 2026-02-28" }, "trust": { "sourceLabel": "tencent", "provenanceUrl": "https://clawhub.ai/cosperypf/terminal-killer", "publisherUrl": "https://clawhub.ai/cosperypf/terminal-killer", "owner": "cosperypf", "version": "1.2.0", "license": null, "verificationStatus": "Indexed source record" }, "links": { "detailUrl": "https://openagent3.xyz/skills/terminal-killer", "downloadUrl": "https://openagent3.xyz/downloads/terminal-killer", "agentUrl": "https://openagent3.xyz/skills/terminal-killer/agent", "manifestUrl": "https://openagent3.xyz/skills/terminal-killer/agent.json", "briefUrl": "https://openagent3.xyz/skills/terminal-killer/agent.md" } }