Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/consuelohq-kiro-7a40a2
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions: not checked yet.
What's inside
- Skill name
kiro- Description
- Orchestrate Kiro CLI coding sessions via ACP. Spawns Kiro for coding tasks, monitors progress, creates branches and PRs when done. Handles both chat-initiated and Linear-spec tasks. Use when asked to code, build, or implement features.
- License
- MIT
- Instructions
- 332 lines in SKILL.md · view on GitHub
Before installing
- Ships executable scripts
Agent Skills spec issues
- body: contains an absolute path from the author's machine
Files (12)
- .specs-watcher-state.json
- NEXT-STEPS.md
- SKILL.md
- com.kiro.webhook-receiver.plist
- config.sh
- kiro_agent.py
- linear-api-kiro.sh
- linear_helper.py
- pipeline-health.sh
- pipeline-status.sh
- specs-watcher.sh
- webhook_receiver.py
SKILL.md (first lines)
# Kiro Agent — Automated Coding via Kiro CLI
You (OpenCode) are the **orchestrator**. Kiro is the **coder**. You spawn Kiro, give it tasks, monitor its work, and handle the git/PR workflow when it's done.
**Everything is remote.** Kiro reads and writes code via the GitHub API — no local file changes. Multiple sessions can run simultaneously on different branches without conflicts.
## When to Use This Skill
- User asks to "build", "implement", "code", "create", or "fix" something or /kiro
- A Linear task is ready for development
- A previous deploy failed and the fix needs to be coded
## Architecture
```
OpenCode (you) ──ACP──> Kiro CLI (codes via GitHub API)
│
├── reads files via github-dev scripts
├── writes files via github-dev scripts
├── commits via github-dev scripts
│
OpenCode (you) ──gh CLI──> GitHub (branch, PR, labels)
```
## Paths & Constants
```
KIRO_ACP_CLIENT = "/Users/kokayi/Dev/claude-agent-workflow/.opencode/skills/kiro-acp/client.py"
GITHUB_DEV = "/Users/kokayi/Dev/claude-agent-workflow/.opencode/skills/github-dev/dev"
GITHUB_DEV_DIR = "/Users/kokayi/Dev/claude-agent-workflow/.opencode/skills/github-dev"
LINEAR_HELPER = "/Users/kokayi/Dev/claude-agent-workflow/.opencode/skills/kiro/linear_helper.py"
KIRO_AGENT = "/Users/kokayi/Dev/claude-agent-workflow/.opencode/skills/kiro/kiro_agent.py"
DEFAULT_CWD = "/Users/kokayi/Dev/consuelo_on_call_coaching"
DEFAULT_REPO = "kokayicobb/consuelo_on_call_coaching"
```
## Full Workflow
### Step 1: Receive Task
Two entry points:
**Chat entry** — user says "build X" or "implement feature Y":
- Extract the coding task from the conversation
- Construct a detailed prompt for Kiro (more detail = better results)
**Linear spec entry** — task ID provided (e.g., CON-456):
- Fetch the spec: `python3 {GITHUB_DEV_DIR}/fetch_task.py CON-456`
- Read the saved spec file: `{GITHUB_DEV_DIR}/tasks/CON-456.md`
- The spec contains requirements, acceptance criteria, and context
### Step 2: Create Branch
Start a github-dev session to create the working branch:
```bash
# For Linear tasks:
python3 {GITHUB_DEV} task CON-456
# For chat tasks (no Linear):
python3 {GITHUB_DEV} create-branch "add user authentication flow"
```
This creates a remote branch on GitHub and saves session state. Note the branch name from the output.
### Step 3: Spawn Kiro ACP Session
Use the orchestrator to spawn and manage Kiro:
```python
import sys
sys.path.insert(0, "/Users/kokayi/Dev/claude-agent-workflow/.opencode/skills/kiro")
from kiro_agent import KiroSession
# Start a session
session = KiroSession(
cwd="/Users/kokayi/Dev/consuelo_on_call_coaching",
repo="kokayicobb/consuelo_on_call_coaching",
)
session.start()
# Send the task
result = session.run_task(
task_description="Implement user authentication with JWT tokens",
branch_name="remote/consuelo_on_call_coaching-a3f2--add-user-auth",
linear_spec=None, # or the spec text if from Linear
linear_task_id=None, # or "CON-456" if from Linear
)
# result contains: what Kiro did, whether it created specs or coded, files changed
```
### Step 4: What Kiro Does (You Don't Control This)
Kiro receives a prompt that instructs it to:
1. **Assess complexity** — is this a straightforward coding task or does it need specs?
2. **If complex** — create Linear issues tagged `kiro` for each piece of work, then STOP
3. **If straightforward** — code the solution using github-dev scripts:
- Read files: `python3 {GITHUB_DEV} read src/file.tsx`
- Explore: `python3 {GITHUB_DEV} ls src/` and `python3 {GITHUB_DEV} tree src/`
- Write files: `echo 'content' | python3 {GITHUB_DEV} write src/new-file.tsx`
- Or: `python3 {GITHUB_DEV} write src/file.tsx --content-file /tmp/edit.tsx`
- Commit: `python3 {GITHUB_DEV} commit "feat: add user auth"`
- Can do multiple write+commit cycles
### Step 5: Monitor Progress
While Kiro works, you monitor via streaming. The orchestrator (`kiro_agent.py`) handles this, but key things to watch for:
- **ToolCall events with "linear" mentions** — Kiro is creating specs → session will end without coding
- **ToolCall events with file operations** — Kiro is coding, relay significant updates to chat
- **Kiro saying "created specs" or "created issues"** — end session, report specs
- **TurnEnd** — Kiro's turn is complete, evaluate
Relay significant progress to Ko's chat:
- "Kiro is reading the codebase structure..."
- "Kiro is writing src/components/Auth.tsx..."From consuelohq/openengineer, MIT.
More from consuelohq/openengineer
- 0.474score
brave-instagram-search
consuelohq
Fast Instagram lead generation using Brave Search API.
SkillDFreshness: D★ 7Claude Code +9 - 0.471score
instagram-lead-scraper
consuelohq
SCALED Instagram lead scraper using Agent Browser.
SkillDFreshness: D★ 7Claude Code +9 - 0.469score
agent-browser
consuelohq
Browser automation for opening, navigating, and interacting with web pages using the agent-browser CLI.
SkillDFreshness: D★ 7Claude Code +9 - 0.466score
deep-search
consuelohq
Hybrid search using Brave for discovery + Agent Browser for deep extraction.
SkillDFreshness: D★ 7Claude Code +9 - 0.465score
github-dev
consuelohq
Remote development via GitHub API.
SkillDFreshness: D★ 7Claude Code +9 - 0.465score
kiro
consuelohq
Orchestrate Kiro CLI coding sessions via ACP.
SkillDFreshness: D★ 7Claude Code +9 - 0.463score
exa-search
consuelohq
Search the web using Exa AI instead of Brave (MCP-compatible endpoint)
SkillDFreshness: D★ 7Claude Code +9 - 0.462score
consuelo-test
consuelohq
Test deployed Consuelo changes against Linear task requirements and PR verification criteria.
SkillDFreshness: D★ 7Claude Code +9 - 0.247score
github-review
consuelohq
No description from the publisher yet.
SkillDFreshness: D★ 7Claude Code +9 - 0.247score
kiro-acp
consuelohq
No description from the publisher yet.
SkillDFreshness: D★ 7Claude Code +9 - 0.247score
tailscale-artifacts
consuelohq
No description from the publisher yet.
SkillDFreshness: D★ 7Claude Code +9 - 0.247score
librechat-artifacts
consuelohq
No description from the publisher yet.
SkillDFreshness: D★ 7Claude Code +9
Signals
- Stars
- 7
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Feb 9, 2026 (8 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/consuelohq/openengineer
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
Same skill elsewhere
Copies of this SKILL.md published by others; search shows only one of them.
More from consuelohq
- 0.474score
brave-instagram-search
consuelohq
Fast Instagram lead generation using Brave Search API.
SkillDFreshness: D★ 7Claude Code +9 - 0.471score
instagram-lead-scraper
consuelohq
SCALED Instagram lead scraper using Agent Browser.
SkillDFreshness: D★ 7Claude Code +9 - 0.469score
agent-browser
consuelohq
Browser automation for opening, navigating, and interacting with web pages using the agent-browser CLI.
SkillDFreshness: D★ 7Claude Code +9 - 0.466score
deep-search
consuelohq
Hybrid search using Brave for discovery + Agent Browser for deep extraction.
SkillDFreshness: D★ 7Claude Code +9 - 0.465score
github-dev
consuelohq
Remote development via GitHub API.
SkillDFreshness: D★ 7Claude Code +9 - 0.465score
kiro
consuelohq
Orchestrate Kiro CLI coding sessions via ACP.
SkillDFreshness: D★ 7Claude Code +9
Cite this page
kiro skill. Agent Mart. https://agentmart.dev/skills/consuelohq-kiro-7a40a2 (last updated 2026-10-24)