| name | kiro-cli |
| description | Use when the user asks to run Kiro CLI in non-interactive mode or references Kiro for AI chat, shell translation, or agent management. |
Kiro CLI Skill Guide
Before running Kiro, follow the shared logging and summary protocol in ../_shared/agent-cli.md.
Running a Task
- Verify installation:
command -v kiro-cli
- Select the mode required for the task; default to
--no-interactive mode for non-interactive use. Default to untrusted/read-only-style operation for analysis. For implementation tasks, first use superpowers:using-git-worktrees to create a fresh git worktree, then run Kiro there with trusted tools when approved.
- Always use
the current host's user-question or approval mechanism before using --trust-all-tools or executing generated shell commands.
- Assemble the command with the appropriate options:
--agent <AGENT_NAME> - Use specific agent
--model <MODEL> - Model to use
--no-interactive - Run without user input (required for non-interactive)
-a, --trust-all-tools - Trust all tools without confirmation
--trust-tools <TOOL_NAMES> - Trust specific tools only (comma-separated)
-r, --resume - Resume most recent conversation
- Do not suppress stderr. Capture stdout and stderr into
full.md, and require the prompt to write summary.md.
- Run the command, inspect the full log and summary, and summarize the outcome for the user.
Quick Reference
| Use case | Command pattern |
|---|
| Non-interactive chat | kiro-cli chat "your input" --no-interactive |
| With specific agent | kiro-cli chat --agent <agent> "input" --no-interactive |
| Trust all tools in fresh worktree only | kiro-cli chat "input" --trust-all-tools --no-interactive |
| Trust write tools in fresh worktree only | kiro-cli chat "input" --trust-tools=fs_read,fs_write --no-interactive |
| Resume conversation | kiro-cli chat --resume --no-interactive |
| List models | kiro-cli chat --list-models |
| List sessions | kiro-cli chat --list-sessions |
Tool Trust Options
| Option | Description | Use Case |
|---|
| (none) | Ask for confirmation | Safest, general use |
--trust-tools <tools> | Trust specific tools | Controlled access; write tools require a fresh worktree |
--trust-all-tools | Trust all tools | Full access in a fresh worktree |
Example Commands
kiro-cli chat "Explain this code" --no-interactive
kiro-cli chat --agent builder "Fix the bug" --trust-all-tools --no-interactive
kiro-cli chat --resume --no-interactive
Following Up
- After every
kiro-cli command, immediately use the current host's user-question or approval mechanism to confirm next steps, collect clarifications, or decide whether to resume with --resume.
- For shell translation, always confirm execution with the user before running the command.
- Restate the chosen agent and tool trust options when proposing follow-up actions.
Error Handling
- Stop and report failures whenever
kiro-cli --version or a kiro-cli command exits non-zero; request direction before retrying.
- Use
kiro-cli doctor to debug installation issues if needed.
- Before you use high-impact flags (
--trust-all-tools) ask the user for permission using the current host's user-question or approval mechanism unless it was already given.
- Never execute a generated shell command without explicit user approval.
- Always validate Kiro's output for security vulnerabilities (XSS, injection) before using.