{
  "schemaVersion": "1.0",
  "item": {
    "slug": "openclaw-kirocli-coding-agent",
    "name": "openclaw-kirocli-coding-agent",
    "source": "tencent",
    "type": "skill",
    "category": "开发工具",
    "sourceUrl": "https://clawhub.ai/dandysuper/openclaw-kirocli-coding-agent",
    "canonicalUrl": "https://clawhub.ai/dandysuper/openclaw-kirocli-coding-agent",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadMode": "redirect",
    "downloadUrl": "/downloads/openclaw-kirocli-coding-agent",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=openclaw-kirocli-coding-agent",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "installMethod": "Manual import",
    "extraction": "Extract archive",
    "prerequisites": [
      "OpenClaw"
    ],
    "packageFormat": "ZIP package",
    "includedAssets": [
      "_meta.json",
      "README.md",
      "SKILL.md",
      "README.zh-CN.md"
    ],
    "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",
      "slug": "openclaw-kirocli-coding-agent",
      "status": "healthy",
      "reason": "direct_download_ok",
      "recommendedAction": "download",
      "checkedAt": "2026-05-12T11:28:17.055Z",
      "expiresAt": "2026-05-19T11:28:17.055Z",
      "httpStatus": 200,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=openclaw-kirocli-coding-agent",
      "contentType": "application/zip",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=openclaw-kirocli-coding-agent",
        "contentDisposition": "attachment; filename=\"openclaw-kirocli-coding-agent-1.5.0.zip\"",
        "redirectLocation": null,
        "bodySnippet": null,
        "slug": "openclaw-kirocli-coding-agent"
      },
      "scope": "item",
      "summary": "Item download looks usable.",
      "detail": "Yavira can redirect you to the upstream package for this item.",
      "primaryActionLabel": "Download for OpenClaw",
      "primaryActionHref": "/downloads/openclaw-kirocli-coding-agent"
    },
    "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/openclaw-kirocli-coding-agent",
    "agentPageUrl": "https://openagent3.xyz/skills/openclaw-kirocli-coding-agent/agent",
    "manifestUrl": "https://openagent3.xyz/skills/openclaw-kirocli-coding-agent/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/openclaw-kirocli-coding-agent/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": "Coding Agent",
        "body": "Launch and manage AI coding agents (Codex, Claude Code, Kiro CLI, OpenCode, Pi) from OpenClaw using bash with background process control."
      },
      {
        "title": "PTY Mode Required",
        "body": "Coding agents are interactive terminal applications that need a pseudo-terminal (PTY). Without it, output breaks or the agent hangs.\n\nAlways set pty:true:\n\n# Correct — with PTY\nbash pty:true command:\"codex exec 'Your prompt'\"\n\n# Wrong — agent may break or hang\nbash command:\"codex exec 'Your prompt'\""
      },
      {
        "title": "Bash Parameters",
        "body": "ParameterTypeDescriptioncommandstringShell command to runptybooleanAllocate a pseudo-terminal (required for coding agents)workdirstringWorking directory (agent sees only this folder)backgroundbooleanRun in background; returns sessionId for monitoringtimeoutnumberTimeout in seconds (kills process on expiry)elevatedbooleanRun on host instead of sandbox (if allowed)"
      },
      {
        "title": "Process Actions (background sessions)",
        "body": "ActionDescriptionlistList all running/recent sessionspollCheck if a session is still runninglogGet session output (optional offset/limit)writeSend raw data to stdin (no newline)submitSend data + newline (like typing and pressing Enter)send-keysSend key tokens or hex bytespastePaste text (optional bracketed mode)killTerminate the session"
      },
      {
        "title": "Quick Start: One-Shot Tasks",
        "body": "# Codex needs a git repo — create a temp one for scratch work\nSCRATCH=$(mktemp -d) && cd $SCRATCH && git init && codex exec \"Your prompt\"\n\n# Run inside an existing project (with PTY)\nbash pty:true workdir:~/project command:\"codex exec 'Add error handling to the API calls'\""
      },
      {
        "title": "Core Pattern: workdir + background + pty",
        "body": "For longer tasks, combine all three:\n\n# 1. Start the agent\nbash pty:true workdir:~/project background:true command:\"codex exec --full-auto 'Build a snake game'\"\n# → returns sessionId\n\n# 2. Monitor progress\nprocess action:log sessionId:XXX\n\n# 3. Check completion\nprocess action:poll sessionId:XXX\n\n# 4. Send input if the agent asks a question\nprocess action:submit sessionId:XXX data:\"yes\"     # text + Enter\nprocess action:write sessionId:XXX data:\"y\"         # raw keystroke\n\n# 5. Kill if stuck\nprocess action:kill sessionId:XXX\n\nWhy workdir? The agent starts in a focused directory and won't wander into unrelated files."
      },
      {
        "title": "Codex CLI",
        "body": "Default model: gpt-5.2-codex (set in ~/.codex/config.toml)"
      },
      {
        "title": "Flags",
        "body": "FlagEffectexec \"prompt\"One-shot execution, exits when done--full-autoSandboxed, auto-approves within workspace--yoloNo sandbox, no approvals (fastest, most dangerous)"
      },
      {
        "title": "Building / Creating",
        "body": "# One-shot with auto-approve\nbash pty:true workdir:~/project command:\"codex exec --full-auto 'Build a dark mode toggle'\"\n\n# Background for longer work\nbash pty:true workdir:~/project background:true command:\"codex --yolo 'Refactor the auth module'\""
      },
      {
        "title": "Reviewing PRs",
        "body": "Never review PRs in OpenClaw's own project folder. Clone to a temp directory or use a git worktree.\n\n# Clone to temp\nREVIEW_DIR=$(mktemp -d)\ngit clone https://github.com/user/repo.git $REVIEW_DIR\ncd $REVIEW_DIR && gh pr checkout 130\nbash pty:true workdir:$REVIEW_DIR command:\"codex review --base origin/main\"\n\n# Or use git worktree (keeps main intact)\ngit worktree add /tmp/pr-130-review pr-130-branch\nbash pty:true workdir:/tmp/pr-130-review command:\"codex review --base main\""
      },
      {
        "title": "Batch PR Reviews",
        "body": "# Fetch all PR refs\ngit fetch origin '+refs/pull/*/head:refs/remotes/origin/pr/*'\n\n# Launch one Codex per PR (all background + PTY)\nbash pty:true workdir:~/project background:true command:\"codex exec 'Review PR #86. git diff origin/main...origin/pr/86'\"\nbash pty:true workdir:~/project background:true command:\"codex exec 'Review PR #87. git diff origin/main...origin/pr/87'\"\n\n# Monitor\nprocess action:list\n\n# Post results\ngh pr comment <PR#> --body \"<review content>\""
      },
      {
        "title": "Claude Code",
        "body": "# One-shot\nbash pty:true workdir:~/project command:\"claude 'Your task'\"\n\n# Background\nbash pty:true workdir:~/project background:true command:\"claude 'Your task'\""
      },
      {
        "title": "Kiro CLI (AWS)",
        "body": "AWS AI coding assistant with session persistence, custom agents, skills, hooks, steering, subagents, planning mode, and MCP integration.\n\nInstall: https://kiro.dev/docs/cli/installation"
      },
      {
        "title": "Basic Usage",
        "body": "kiro-cli                           # Interactive chat (default)\nkiro-cli chat \"Your question\"      # Direct question\nkiro-cli --agent my-agent          # Use a specific agent\nkiro-cli chat --resume             # Resume last session (per-directory)\nkiro-cli chat --resume-picker      # Pick from saved sessions\nkiro-cli chat --list-sessions      # List all sessions"
      },
      {
        "title": "Non-Interactive Mode",
        "body": "For scripting and automation — outputs a single response to STDOUT, then exits.\n\n# Single response\nkiro-cli chat --no-interactive \"Show current directory\"\n\n# Trust all tools (no confirmation prompts)\nkiro-cli chat --no-interactive --trust-all-tools \"Create hello.py\"\n\n# Trust specific tools only\nkiro-cli chat --no-interactive --trust-tools \"fs_read,fs_write\" \"Read package.json\"\n\nTool trust: --trust-all-tools for full automation. For untrusted input, use --trust-tools \"fs_read,fs_write,shell\" to limit scope."
      },
      {
        "title": "OpenClaw Integration",
        "body": "# Interactive session (background)\nbash pty:true workdir:~/project background:true command:\"kiro-cli\"\n\n# One-shot query (non-interactive)\nbash pty:true workdir:~/project command:\"kiro-cli chat --no-interactive --trust-all-tools 'List all TODO comments in src/'\"\n\n# With a specific agent\nbash pty:true workdir:~/project background:true command:\"kiro-cli --agent aws-expert 'Set up Lambda'\"\n\n# Resume previous session\nbash pty:true workdir:~/project command:\"kiro-cli chat --resume\""
      },
      {
        "title": "Skills (Agent Skills)",
        "body": "Skills are portable instruction packages that extend what Kiro knows. When a request matches a skill's description, Kiro automatically loads and follows its instructions — no slash command needed.\n\nSkill Locations\n\nLocationScopeNotes.kiro/skills/<name>/Workspace (project)Shared via version control~/.kiro/skills/<name>/Global (all projects)Personal workflows\n\nWorkspace skills take priority when names collide.\n\nCreating a Skill\n\nA skill is a folder with a SKILL.md file:\n\nmy-skill/\n├── SKILL.md          # Required — frontmatter + instructions\n└── references/       # Optional — detailed docs loaded on demand\n    └── guide.md\n\nSKILL.md format:\n\n---\nname: pr-review\ndescription: Review pull requests for code quality, security issues, and test coverage.\n---\n\n## Review checklist\n\n1. Check for vulnerabilities, injection risks, exposed secrets\n2. Verify edge cases and failure modes are handled\n3. Confirm new code has appropriate tests\n\nFor detailed patterns, see `references/guide.md`.\n\nname — Unique identifier for the skill.\ndescription — Determines when Kiro activates the skill. Be specific; include keywords that match how you'd phrase requests.\nReference files — Stored in references/. Kiro loads them only when the instructions direct it to.\n\nSkills in Custom Agents\n\nThe default agent auto-discovers skills. Custom agents need explicit resource declarations:\n\n{\n  \"name\": \"my-agent\",\n  \"resources\": [\n    \"skill://.kiro/skills/*/SKILL.md\",\n    \"skill://~/.kiro/skills/*/SKILL.md\"\n  ]\n}\n\nSkill Best Practices\n\nPrecise descriptions — \"Review pull requests for security vulnerabilities and test coverage\" activates reliably; \"Helps with code review\" does not.\nKeep SKILL.md actionable — Put lengthy reference material in references/ files.\nRight scope — Global skills for personal workflows; workspace skills for team/project conventions.\nVersion control — Commit .kiro/skills/ so the team shares the same workflows.\nCheck availability — Use /context show to see which skills are loaded in the current session."
      },
      {
        "title": "Planning Mode (Plan Agent)",
        "body": "Plan Agent is a built-in read-only agent for structured planning before execution. It transforms ideas into detailed implementation plans through an interactive workflow.\n\nWhen to Use\n\nComplex multi-step features (e.g., \"build a user authentication system\")\nUnclear or evolving requirements that need refinement\nLarge features that benefit from task breakdown before coding\n\nWhen NOT to Use\n\nSimple queries or single-step tasks\nUser already has clear, specific instructions\nQuick fixes or small changes\n\nHow to Enter Planning Mode\n\n# Slash command\n> /plan\n\n# With an immediate prompt\n> /plan Build a REST API for user authentication\n\n# Keyboard shortcut (toggles plan ↔ execution)\nShift + Tab\n\nWhen active, the prompt shows a [plan] indicator.\n\nPlan Workflow (4 phases)\n\nRequirements gathering — Structured multiple-choice questions to refine your idea. Answer with 1=a, 2=b syntax or free-text.\nResearch & analysis — Explores your codebase using code intelligence, grep, and glob tools.\nImplementation plan — Produces a task breakdown with clear objectives, implementation guidance, and demo descriptions for each task.\nApproval & handoff — You review the plan. On approval (y), the plan transfers automatically to the execution agent.\n\nPlan Agent is read-only: it can read files, search code, and research, but cannot write files or execute commands until handoff.\n\nOpenClaw Integration for Planning Mode\n\nFor interactive planning sessions, run Kiro in background mode and relay the /plan command:\n\n# Start interactive Kiro session\nbash pty:true workdir:~/project background:true command:\"kiro-cli chat --trust-all-tools\"\n\n# Enter planning mode\nprocess action:submit sessionId:XXX data:\"/plan Build a REST API for user authentication\"\n\n# Relay the user's answers to requirement questions\nprocess action:submit sessionId:XXX data:\"1=a, 2=d I'm using Rust with Axum\"\n\n# Approve the plan\nprocess action:submit sessionId:XXX data:\"y\"\n\n# Monitor output\nprocess action:log sessionId:XXX\n\nExample Planning Session\n\n> /plan Add user authentication to my web app\n\n[plan] > I understand you want to add user authentication.\n[1]: What authentication method?\n  a. Email/Password   b. OAuth   c. Magic Links   d. Multi-factor\n> 1=a\n\n[plan] > Great! Email/password it is.\n[2]: What's your tech stack?\n  a. React + Node.js   b. Next.js   c. Django/Flask   d. Other\n> 2=d, I'm using Rust with Axum\n\n[plan] > Researching Axum authentication patterns...\n\n**Implementation Plan — User Authentication System**\n[Detailed task breakdown...]\n\nDoes this plan look good? Ready to exit [plan] agent? [y/n]: y\n[default] > Implement this plan: [Plan transferred]"
      },
      {
        "title": "Hooks",
        "body": "Hooks execute custom commands at specific points during agent lifecycle and tool execution. Defined in the agent configuration file.\n\nHook Types\n\nHookTriggerCan Block?AgentSpawnAgent startsNoUserPromptSubmitUser sends a promptNoPreToolUseBefore a tool runsYes (exit code 2)PostToolUseAfter a tool runsNoStopAgent finishes a turnNo\n\nExit Codes\n\n0 — Success. STDOUT captured (added to context for AgentSpawn/UserPromptSubmit).\n2 — (PreToolUse only) Block tool execution; STDERR returned to the LLM.\nOther — Failure. STDERR shown as warning to user.\n\nTool Matching\n\nUse the matcher field to target specific tools:\n\nMatcherMatches\"fs_write\" or \"write\"Write tool\"execute_bash\" or \"shell\"Shell execution\"@git\"All tools from git MCP server\"@git/status\"Specific MCP tool\"*\"All tools (built-in + MCP)\"@builtin\"Built-in tools only\n\nConfiguration\n\ntimeout_ms — Default 30,000ms (30s).\ncache_ttl_seconds — 0 = no caching (default); > 0 = cache successful results. AgentSpawn hooks are never cached.\n\nSee Agent Configuration Reference for full syntax."
      },
      {
        "title": "Subagents",
        "body": "Kiro can delegate tasks to subagents — independent agents with their own context that run autonomously and return results.\n\n> Use the backend agent to refactor the payment module\n\nKey capabilities:\n\nAutonomous execution with isolated context\nLive progress tracking\nParallel execution for multiple tasks\nCustom agent configurations for specialized workflows\n\nAvailable tools in subagents: read, write, shell, code intelligence, MCP tools.\nNot available: web_search, web_fetch, use_aws, grep, glob, thinking, todo_list."
      },
      {
        "title": "Custom Agents",
        "body": "Pre-define tool permissions, context resources, and behaviors:\n\nkiro-cli agent list              # List available agents\nkiro-cli agent create my-agent   # Create new agent\nkiro-cli agent edit my-agent     # Edit agent config\nkiro-cli agent validate ./a.json # Validate config file\nkiro-cli agent set-default my-agent\n\nBenefits: Pre-approved tool trust, limited tool access, auto-loaded project docs, shareable team configs."
      },
      {
        "title": "Steering (Project Context)",
        "body": "Provide persistent project knowledge via markdown files:\n\nPathScope.kiro/steering/Workspace — this project only~/.kiro/steering/Global — all projects\n\nExample structure:\n\n.kiro/steering/\n├── product.md           # Product overview\n├── tech.md              # Tech stack\n├── structure.md         # Project structure\n└── api-standards.md     # API conventions\n\nAlso supports AGENTS.md in the project root or ~/.kiro/steering/.\n\nIn custom agents: Add \"resources\": [\"file://.kiro/steering/**/*.md\"] to config."
      },
      {
        "title": "MCP Integration",
        "body": "Connect external tools and data sources via Model Context Protocol:\n\nkiro-cli mcp add --name my-server --command \"node server.js\" --scope workspace\nkiro-cli mcp list [workspace|global]\nkiro-cli mcp status --name my-server\nkiro-cli mcp remove --name my-server --scope workspace"
      },
      {
        "title": "OpenCode",
        "body": "bash pty:true workdir:~/project command:\"opencode run 'Your task'\""
      },
      {
        "title": "Pi Coding Agent",
        "body": "# Install: npm install -g @mariozechner/pi-coding-agent\n\n# Interactive\nbash pty:true workdir:~/project command:\"pi 'Your task'\"\n\n# Non-interactive (single response)\nbash pty:true command:\"pi -p 'Summarize src/'\"\n\n# Different provider/model\nbash pty:true command:\"pi --provider openai --model gpt-4o-mini -p 'Your task'\""
      },
      {
        "title": "Parallel Issue Fixing (git worktrees)",
        "body": "Fix multiple issues simultaneously using isolated worktrees:\n\n# 1. Create worktrees\ngit worktree add -b fix/issue-78 /tmp/issue-78 main\ngit worktree add -b fix/issue-99 /tmp/issue-99 main\n\n# 2. Launch agents (background + PTY)\nbash pty:true workdir:/tmp/issue-78 background:true command:\"pnpm install && codex --yolo 'Fix issue #78: <description>. Commit and push.'\"\nbash pty:true workdir:/tmp/issue-99 background:true command:\"pnpm install && codex --yolo 'Fix issue #99: <description>. Commit and push.'\"\n\n# 3. Monitor\nprocess action:list\nprocess action:log sessionId:XXX\n\n# 4. Create PRs\ncd /tmp/issue-78 && git push -u origin fix/issue-78\ngh pr create --repo user/repo --head fix/issue-78 --title \"fix: ...\" --body \"...\"\n\n# 5. Clean up\ngit worktree remove /tmp/issue-78\ngit worktree remove /tmp/issue-99"
      },
      {
        "title": "Rules",
        "body": "Always use pty:true — Coding agents need a pseudo-terminal.\nRespect tool choice — If the user asks for Kiro, use Kiro; for Codex, use Codex. Do NOT hand-code patches yourself when orchestrating agents. If an agent fails or hangs, respawn it or ask the user — don't silently take over.\nBe patient — Don't kill sessions just because they seem slow.\nMonitor with process:log — Check progress without interfering.\nCodex auto-approve flags — Use --full-auto (sandboxed) or --yolo (no sandbox) for building tasks.\nKiro tool trust — Use --trust-all-tools for automation; --trust-tools for restricted scope.\nKiro one-shots — Use --no-interactive for single-response queries.\nParallel is OK — Run multiple agent processes concurrently for batch work.\nNever start agents in ~/clawd/ — Agents will read system docs and behave unpredictably.\nNever checkout branches in ~/Projects/openclaw/ — That's the live OpenClaw instance.\nSuggest Kiro /plan for complex tasks — When requirements are unclear or multi-step, suggest Plan Agent and let the user decide.\nLeverage Kiro skills — If the project has .kiro/skills/ or the user has ~/.kiro/skills/, relevant skills activate automatically. No extra flags needed."
      },
      {
        "title": "Progress Updates",
        "body": "When you spawn coding agents in the background, keep the user informed:\n\nOn start — 1 short message: what's running, where, and which agent.\nOn change — Update only when something happens:\n\nA milestone completes (build finished, tests passed)\nThe agent asks a question or needs input\nAn error occurs or user action is needed\nThe agent finishes (include what changed and where)\n\n\nOn kill — Immediately say you killed it and why."
      },
      {
        "title": "Auto-Notify on Completion",
        "body": "For long-running tasks, append a wake trigger so OpenClaw is notified immediately when the agent finishes:\n\n... your task here.\n\nWhen completely finished, run this command to notify me:\nopenclaw gateway wake --text \"Done: [brief summary]\" --mode now\n\nExample:\n\nbash pty:true workdir:~/project background:true command:\"codex --yolo exec 'Build a REST API for todos.\n\nWhen completely finished, run: openclaw gateway wake --text \\\"Done: Built todos REST API with CRUD endpoints\\\" --mode now'\"\n\nThis triggers an immediate wake event instead of waiting for the next heartbeat."
      },
      {
        "title": "Learnings",
        "body": "PTY is essential — Without pty:true, output breaks or the agent hangs.\nGit repo required for Codex — Use mktemp -d && git init for scratch work.\nexec for one-shots — codex exec \"prompt\" runs and exits cleanly.\nsubmit vs write — submit sends input + Enter; write sends raw data without newline.\nSkills activate automatically — No slash command needed; Kiro matches your request against skill descriptions.\nPlan before complex builds — /plan saves time on multi-step features by clarifying requirements upfront."
      }
    ],
    "body": "Coding Agent\n\nLaunch and manage AI coding agents (Codex, Claude Code, Kiro CLI, OpenCode, Pi) from OpenClaw using bash with background process control.\n\nPTY Mode Required\n\nCoding agents are interactive terminal applications that need a pseudo-terminal (PTY). Without it, output breaks or the agent hangs.\n\nAlways set pty:true:\n\n# Correct — with PTY\nbash pty:true command:\"codex exec 'Your prompt'\"\n\n# Wrong — agent may break or hang\nbash command:\"codex exec 'Your prompt'\"\n\nTool Reference\nBash Parameters\nParameter\tType\tDescription\ncommand\tstring\tShell command to run\npty\tboolean\tAllocate a pseudo-terminal (required for coding agents)\nworkdir\tstring\tWorking directory (agent sees only this folder)\nbackground\tboolean\tRun in background; returns sessionId for monitoring\ntimeout\tnumber\tTimeout in seconds (kills process on expiry)\nelevated\tboolean\tRun on host instead of sandbox (if allowed)\nProcess Actions (background sessions)\nAction\tDescription\nlist\tList all running/recent sessions\npoll\tCheck if a session is still running\nlog\tGet session output (optional offset/limit)\nwrite\tSend raw data to stdin (no newline)\nsubmit\tSend data + newline (like typing and pressing Enter)\nsend-keys\tSend key tokens or hex bytes\npaste\tPaste text (optional bracketed mode)\nkill\tTerminate the session\nQuick Start: One-Shot Tasks\n# Codex needs a git repo — create a temp one for scratch work\nSCRATCH=$(mktemp -d) && cd $SCRATCH && git init && codex exec \"Your prompt\"\n\n# Run inside an existing project (with PTY)\nbash pty:true workdir:~/project command:\"codex exec 'Add error handling to the API calls'\"\n\nCore Pattern: workdir + background + pty\n\nFor longer tasks, combine all three:\n\n# 1. Start the agent\nbash pty:true workdir:~/project background:true command:\"codex exec --full-auto 'Build a snake game'\"\n# → returns sessionId\n\n# 2. Monitor progress\nprocess action:log sessionId:XXX\n\n# 3. Check completion\nprocess action:poll sessionId:XXX\n\n# 4. Send input if the agent asks a question\nprocess action:submit sessionId:XXX data:\"yes\"     # text + Enter\nprocess action:write sessionId:XXX data:\"y\"         # raw keystroke\n\n# 5. Kill if stuck\nprocess action:kill sessionId:XXX\n\n\nWhy workdir? The agent starts in a focused directory and won't wander into unrelated files.\n\nCodex CLI\n\nDefault model: gpt-5.2-codex (set in ~/.codex/config.toml)\n\nFlags\nFlag\tEffect\nexec \"prompt\"\tOne-shot execution, exits when done\n--full-auto\tSandboxed, auto-approves within workspace\n--yolo\tNo sandbox, no approvals (fastest, most dangerous)\nBuilding / Creating\n# One-shot with auto-approve\nbash pty:true workdir:~/project command:\"codex exec --full-auto 'Build a dark mode toggle'\"\n\n# Background for longer work\nbash pty:true workdir:~/project background:true command:\"codex --yolo 'Refactor the auth module'\"\n\nReviewing PRs\n\nNever review PRs in OpenClaw's own project folder. Clone to a temp directory or use a git worktree.\n\n# Clone to temp\nREVIEW_DIR=$(mktemp -d)\ngit clone https://github.com/user/repo.git $REVIEW_DIR\ncd $REVIEW_DIR && gh pr checkout 130\nbash pty:true workdir:$REVIEW_DIR command:\"codex review --base origin/main\"\n\n# Or use git worktree (keeps main intact)\ngit worktree add /tmp/pr-130-review pr-130-branch\nbash pty:true workdir:/tmp/pr-130-review command:\"codex review --base main\"\n\nBatch PR Reviews\n# Fetch all PR refs\ngit fetch origin '+refs/pull/*/head:refs/remotes/origin/pr/*'\n\n# Launch one Codex per PR (all background + PTY)\nbash pty:true workdir:~/project background:true command:\"codex exec 'Review PR #86. git diff origin/main...origin/pr/86'\"\nbash pty:true workdir:~/project background:true command:\"codex exec 'Review PR #87. git diff origin/main...origin/pr/87'\"\n\n# Monitor\nprocess action:list\n\n# Post results\ngh pr comment <PR#> --body \"<review content>\"\n\nClaude Code\n# One-shot\nbash pty:true workdir:~/project command:\"claude 'Your task'\"\n\n# Background\nbash pty:true workdir:~/project background:true command:\"claude 'Your task'\"\n\nKiro CLI (AWS)\n\nAWS AI coding assistant with session persistence, custom agents, skills, hooks, steering, subagents, planning mode, and MCP integration.\n\nInstall: https://kiro.dev/docs/cli/installation\n\nBasic Usage\nkiro-cli                           # Interactive chat (default)\nkiro-cli chat \"Your question\"      # Direct question\nkiro-cli --agent my-agent          # Use a specific agent\nkiro-cli chat --resume             # Resume last session (per-directory)\nkiro-cli chat --resume-picker      # Pick from saved sessions\nkiro-cli chat --list-sessions      # List all sessions\n\nNon-Interactive Mode\n\nFor scripting and automation — outputs a single response to STDOUT, then exits.\n\n# Single response\nkiro-cli chat --no-interactive \"Show current directory\"\n\n# Trust all tools (no confirmation prompts)\nkiro-cli chat --no-interactive --trust-all-tools \"Create hello.py\"\n\n# Trust specific tools only\nkiro-cli chat --no-interactive --trust-tools \"fs_read,fs_write\" \"Read package.json\"\n\n\nTool trust: --trust-all-tools for full automation. For untrusted input, use --trust-tools \"fs_read,fs_write,shell\" to limit scope.\n\nOpenClaw Integration\n# Interactive session (background)\nbash pty:true workdir:~/project background:true command:\"kiro-cli\"\n\n# One-shot query (non-interactive)\nbash pty:true workdir:~/project command:\"kiro-cli chat --no-interactive --trust-all-tools 'List all TODO comments in src/'\"\n\n# With a specific agent\nbash pty:true workdir:~/project background:true command:\"kiro-cli --agent aws-expert 'Set up Lambda'\"\n\n# Resume previous session\nbash pty:true workdir:~/project command:\"kiro-cli chat --resume\"\n\nSkills (Agent Skills)\n\nSkills are portable instruction packages that extend what Kiro knows. When a request matches a skill's description, Kiro automatically loads and follows its instructions — no slash command needed.\n\nSkill Locations\nLocation\tScope\tNotes\n.kiro/skills/<name>/\tWorkspace (project)\tShared via version control\n~/.kiro/skills/<name>/\tGlobal (all projects)\tPersonal workflows\n\nWorkspace skills take priority when names collide.\n\nCreating a Skill\n\nA skill is a folder with a SKILL.md file:\n\nmy-skill/\n├── SKILL.md          # Required — frontmatter + instructions\n└── references/       # Optional — detailed docs loaded on demand\n    └── guide.md\n\n\nSKILL.md format:\n\n---\nname: pr-review\ndescription: Review pull requests for code quality, security issues, and test coverage.\n---\n\n## Review checklist\n\n1. Check for vulnerabilities, injection risks, exposed secrets\n2. Verify edge cases and failure modes are handled\n3. Confirm new code has appropriate tests\n\nFor detailed patterns, see `references/guide.md`.\n\nname — Unique identifier for the skill.\ndescription — Determines when Kiro activates the skill. Be specific; include keywords that match how you'd phrase requests.\nReference files — Stored in references/. Kiro loads them only when the instructions direct it to.\nSkills in Custom Agents\n\nThe default agent auto-discovers skills. Custom agents need explicit resource declarations:\n\n{\n  \"name\": \"my-agent\",\n  \"resources\": [\n    \"skill://.kiro/skills/*/SKILL.md\",\n    \"skill://~/.kiro/skills/*/SKILL.md\"\n  ]\n}\n\nSkill Best Practices\nPrecise descriptions — \"Review pull requests for security vulnerabilities and test coverage\" activates reliably; \"Helps with code review\" does not.\nKeep SKILL.md actionable — Put lengthy reference material in references/ files.\nRight scope — Global skills for personal workflows; workspace skills for team/project conventions.\nVersion control — Commit .kiro/skills/ so the team shares the same workflows.\nCheck availability — Use /context show to see which skills are loaded in the current session.\nPlanning Mode (Plan Agent)\n\nPlan Agent is a built-in read-only agent for structured planning before execution. It transforms ideas into detailed implementation plans through an interactive workflow.\n\nWhen to Use\nComplex multi-step features (e.g., \"build a user authentication system\")\nUnclear or evolving requirements that need refinement\nLarge features that benefit from task breakdown before coding\nWhen NOT to Use\nSimple queries or single-step tasks\nUser already has clear, specific instructions\nQuick fixes or small changes\nHow to Enter Planning Mode\n# Slash command\n> /plan\n\n# With an immediate prompt\n> /plan Build a REST API for user authentication\n\n# Keyboard shortcut (toggles plan ↔ execution)\nShift + Tab\n\n\nWhen active, the prompt shows a [plan] indicator.\n\nPlan Workflow (4 phases)\nRequirements gathering — Structured multiple-choice questions to refine your idea. Answer with 1=a, 2=b syntax or free-text.\nResearch & analysis — Explores your codebase using code intelligence, grep, and glob tools.\nImplementation plan — Produces a task breakdown with clear objectives, implementation guidance, and demo descriptions for each task.\nApproval & handoff — You review the plan. On approval (y), the plan transfers automatically to the execution agent.\n\nPlan Agent is read-only: it can read files, search code, and research, but cannot write files or execute commands until handoff.\n\nOpenClaw Integration for Planning Mode\n\nFor interactive planning sessions, run Kiro in background mode and relay the /plan command:\n\n# Start interactive Kiro session\nbash pty:true workdir:~/project background:true command:\"kiro-cli chat --trust-all-tools\"\n\n# Enter planning mode\nprocess action:submit sessionId:XXX data:\"/plan Build a REST API for user authentication\"\n\n# Relay the user's answers to requirement questions\nprocess action:submit sessionId:XXX data:\"1=a, 2=d I'm using Rust with Axum\"\n\n# Approve the plan\nprocess action:submit sessionId:XXX data:\"y\"\n\n# Monitor output\nprocess action:log sessionId:XXX\n\nExample Planning Session\n> /plan Add user authentication to my web app\n\n[plan] > I understand you want to add user authentication.\n[1]: What authentication method?\n  a. Email/Password   b. OAuth   c. Magic Links   d. Multi-factor\n> 1=a\n\n[plan] > Great! Email/password it is.\n[2]: What's your tech stack?\n  a. React + Node.js   b. Next.js   c. Django/Flask   d. Other\n> 2=d, I'm using Rust with Axum\n\n[plan] > Researching Axum authentication patterns...\n\n**Implementation Plan — User Authentication System**\n[Detailed task breakdown...]\n\nDoes this plan look good? Ready to exit [plan] agent? [y/n]: y\n[default] > Implement this plan: [Plan transferred]\n\nHooks\n\nHooks execute custom commands at specific points during agent lifecycle and tool execution. Defined in the agent configuration file.\n\nHook Types\nHook\tTrigger\tCan Block?\nAgentSpawn\tAgent starts\tNo\nUserPromptSubmit\tUser sends a prompt\tNo\nPreToolUse\tBefore a tool runs\tYes (exit code 2)\nPostToolUse\tAfter a tool runs\tNo\nStop\tAgent finishes a turn\tNo\nExit Codes\n0 — Success. STDOUT captured (added to context for AgentSpawn/UserPromptSubmit).\n2 — (PreToolUse only) Block tool execution; STDERR returned to the LLM.\nOther — Failure. STDERR shown as warning to user.\nTool Matching\n\nUse the matcher field to target specific tools:\n\nMatcher\tMatches\n\"fs_write\" or \"write\"\tWrite tool\n\"execute_bash\" or \"shell\"\tShell execution\n\"@git\"\tAll tools from git MCP server\n\"@git/status\"\tSpecific MCP tool\n\"*\"\tAll tools (built-in + MCP)\n\"@builtin\"\tBuilt-in tools only\nConfiguration\ntimeout_ms — Default 30,000ms (30s).\ncache_ttl_seconds — 0 = no caching (default); > 0 = cache successful results. AgentSpawn hooks are never cached.\n\nSee Agent Configuration Reference for full syntax.\n\nSubagents\n\nKiro can delegate tasks to subagents — independent agents with their own context that run autonomously and return results.\n\n> Use the backend agent to refactor the payment module\n\n\nKey capabilities:\n\nAutonomous execution with isolated context\nLive progress tracking\nParallel execution for multiple tasks\nCustom agent configurations for specialized workflows\n\nAvailable tools in subagents: read, write, shell, code intelligence, MCP tools. Not available: web_search, web_fetch, use_aws, grep, glob, thinking, todo_list.\n\nCustom Agents\n\nPre-define tool permissions, context resources, and behaviors:\n\nkiro-cli agent list              # List available agents\nkiro-cli agent create my-agent   # Create new agent\nkiro-cli agent edit my-agent     # Edit agent config\nkiro-cli agent validate ./a.json # Validate config file\nkiro-cli agent set-default my-agent\n\n\nBenefits: Pre-approved tool trust, limited tool access, auto-loaded project docs, shareable team configs.\n\nSteering (Project Context)\n\nProvide persistent project knowledge via markdown files:\n\nPath\tScope\n.kiro/steering/\tWorkspace — this project only\n~/.kiro/steering/\tGlobal — all projects\n\nExample structure:\n\n.kiro/steering/\n├── product.md           # Product overview\n├── tech.md              # Tech stack\n├── structure.md         # Project structure\n└── api-standards.md     # API conventions\n\n\nAlso supports AGENTS.md in the project root or ~/.kiro/steering/.\n\nIn custom agents: Add \"resources\": [\"file://.kiro/steering/**/*.md\"] to config.\n\nMCP Integration\n\nConnect external tools and data sources via Model Context Protocol:\n\nkiro-cli mcp add --name my-server --command \"node server.js\" --scope workspace\nkiro-cli mcp list [workspace|global]\nkiro-cli mcp status --name my-server\nkiro-cli mcp remove --name my-server --scope workspace\n\nOpenCode\nbash pty:true workdir:~/project command:\"opencode run 'Your task'\"\n\nPi Coding Agent\n# Install: npm install -g @mariozechner/pi-coding-agent\n\n# Interactive\nbash pty:true workdir:~/project command:\"pi 'Your task'\"\n\n# Non-interactive (single response)\nbash pty:true command:\"pi -p 'Summarize src/'\"\n\n# Different provider/model\nbash pty:true command:\"pi --provider openai --model gpt-4o-mini -p 'Your task'\"\n\nParallel Issue Fixing (git worktrees)\n\nFix multiple issues simultaneously using isolated worktrees:\n\n# 1. Create worktrees\ngit worktree add -b fix/issue-78 /tmp/issue-78 main\ngit worktree add -b fix/issue-99 /tmp/issue-99 main\n\n# 2. Launch agents (background + PTY)\nbash pty:true workdir:/tmp/issue-78 background:true command:\"pnpm install && codex --yolo 'Fix issue #78: <description>. Commit and push.'\"\nbash pty:true workdir:/tmp/issue-99 background:true command:\"pnpm install && codex --yolo 'Fix issue #99: <description>. Commit and push.'\"\n\n# 3. Monitor\nprocess action:list\nprocess action:log sessionId:XXX\n\n# 4. Create PRs\ncd /tmp/issue-78 && git push -u origin fix/issue-78\ngh pr create --repo user/repo --head fix/issue-78 --title \"fix: ...\" --body \"...\"\n\n# 5. Clean up\ngit worktree remove /tmp/issue-78\ngit worktree remove /tmp/issue-99\n\nRules\nAlways use pty:true — Coding agents need a pseudo-terminal.\nRespect tool choice — If the user asks for Kiro, use Kiro; for Codex, use Codex. Do NOT hand-code patches yourself when orchestrating agents. If an agent fails or hangs, respawn it or ask the user — don't silently take over.\nBe patient — Don't kill sessions just because they seem slow.\nMonitor with process:log — Check progress without interfering.\nCodex auto-approve flags — Use --full-auto (sandboxed) or --yolo (no sandbox) for building tasks.\nKiro tool trust — Use --trust-all-tools for automation; --trust-tools for restricted scope.\nKiro one-shots — Use --no-interactive for single-response queries.\nParallel is OK — Run multiple agent processes concurrently for batch work.\nNever start agents in ~/clawd/ — Agents will read system docs and behave unpredictably.\nNever checkout branches in ~/Projects/openclaw/ — That's the live OpenClaw instance.\nSuggest Kiro /plan for complex tasks — When requirements are unclear or multi-step, suggest Plan Agent and let the user decide.\nLeverage Kiro skills — If the project has .kiro/skills/ or the user has ~/.kiro/skills/, relevant skills activate automatically. No extra flags needed.\nProgress Updates\n\nWhen you spawn coding agents in the background, keep the user informed:\n\nOn start — 1 short message: what's running, where, and which agent.\nOn change — Update only when something happens:\nA milestone completes (build finished, tests passed)\nThe agent asks a question or needs input\nAn error occurs or user action is needed\nThe agent finishes (include what changed and where)\nOn kill — Immediately say you killed it and why.\nAuto-Notify on Completion\n\nFor long-running tasks, append a wake trigger so OpenClaw is notified immediately when the agent finishes:\n\n... your task here.\n\nWhen completely finished, run this command to notify me:\nopenclaw gateway wake --text \"Done: [brief summary]\" --mode now\n\n\nExample:\n\nbash pty:true workdir:~/project background:true command:\"codex --yolo exec 'Build a REST API for todos.\n\nWhen completely finished, run: openclaw gateway wake --text \\\"Done: Built todos REST API with CRUD endpoints\\\" --mode now'\"\n\n\nThis triggers an immediate wake event instead of waiting for the next heartbeat.\n\nLearnings\nPTY is essential — Without pty:true, output breaks or the agent hangs.\nGit repo required for Codex — Use mktemp -d && git init for scratch work.\nexec for one-shots — codex exec \"prompt\" runs and exits cleanly.\nsubmit vs write — submit sends input + Enter; write sends raw data without newline.\nSkills activate automatically — No slash command needed; Kiro matches your request against skill descriptions.\nPlan before complex builds — /plan saves time on multi-step features by clarifying requirements upfront."
  },
  "trust": {
    "sourceLabel": "tencent",
    "provenanceUrl": "https://clawhub.ai/dandysuper/openclaw-kirocli-coding-agent",
    "publisherUrl": "https://clawhub.ai/dandysuper/openclaw-kirocli-coding-agent",
    "owner": "dandysuper",
    "version": "1.5.0",
    "license": null,
    "verificationStatus": "Indexed source record"
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/openclaw-kirocli-coding-agent",
    "downloadUrl": "https://openagent3.xyz/downloads/openclaw-kirocli-coding-agent",
    "agentUrl": "https://openagent3.xyz/skills/openclaw-kirocli-coding-agent/agent",
    "manifestUrl": "https://openagent3.xyz/skills/openclaw-kirocli-coding-agent/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/openclaw-kirocli-coding-agent/agent.md"
  }
}