← Back to Skills Marketplace
IDE Agent Kit
by
Petrus Pennanen
· GitHub ↗
· v0.4.1
583
Downloads
0
Stars
1
Active Installs
10
Versions
Install in OpenClaw
/install ide-agent-kit
Description
Filesystem message bus and webhook relay for multi-agent IDE coordination. Use when agents need to share events, poll Ant Farm rooms, receive GitHub/GitLab w...
README (SKILL.md)
IDE Agent Kit
Connect your IDE coding agents into real-time teams through OpenClaw. Filesystem-based message bus, room polling, automation rules, and multi-model agent coordination. Zero dependencies.
Security Model
This skill operates in two tiers:
Core (local-only, no credentials needed):
- Local filesystem queue and receipt log — agents read/write files in the working directory
init,receipt tail,memory(local backend),keepalive— no network, no secretsservebinds to127.0.0.1only by default — receives webhooks, writes to local queue
Advanced (requires explicit opt-in and credentials):
sessions,gateway— talk to an OpenClaw gateway (requiresopenclaw.tokenin config)poll— connects to Ant Farm rooms (requires--api-keyflag)emit,hooks create— POST data to external URLs you specifytmux run,exec— execute shell commands (restricted to an allowlist in config)
No advanced features activate without explicit configuration. The default init config has empty credential fields and a minimal command allowlist.
Network behavior
| Command | Outbound connections | Inbound connections |
|---|---|---|
init, receipt tail, memory (local), keepalive |
None | None |
serve |
None | localhost:8787 only (configurable) |
poll |
Ant Farm API (HTTPS) | None |
sessions, gateway |
OpenClaw gateway (localhost by default) | None |
emit |
User-specified URL | None |
hooks create |
User-specified webhook URL | None |
Command execution
tmux run and exec only run commands listed in tmux.allow in your config. Default allowlist: npm test, npm run build, pytest, git status, git diff. Commands not on the list are rejected.
exec adds an approval flow: commands go through exec request → human/agent exec resolve before running.
Quick Start
npm install -g ide-agent-kit
ide-agent-kit init --ide claude-code
Creates a local ide-agent-kit.json config. All credential fields are blank. Nothing connects to any server until you configure it.
Connectivity Modes
Four modes that compose freely. Only mode 1 is active by default.
1. Local Filesystem Bus (default)
Agents on the same machine communicate through a shared queue directory and receipt log. No network, no server, no API keys.
- Queue:
./ide-agent-queue.jsonl - Receipts:
./ide-agent-receipts.jsonl
2. Webhook Relay Server (optional)
Receives inbound webhooks from GitHub/GitLab and writes them to the local event queue.
ide-agent-kit serve [--config \x3Cpath>]
Binds to 127.0.0.1:8787 by default. Set github.webhook_secret in config to verify signatures. Does not make outbound connections.
3. Ant Farm Room Polling (optional)
Connects to Ant Farm rooms for cross-machine coordination.
ide-agent-kit poll --rooms \x3Croom1,room2> --api-key \x3Ckey> --handle \x3C@handle> [--interval \x3Csec>]
Requires: --api-key flag (Ant Farm API key). Rate-limited, default 120s interval.
4. GitHub Events (optional)
When serve is running, point a GitHub webhook at your relay URL. Translates PR/issue/CI events into local queue events.
Requires: github.webhook_secret in config to verify inbound signatures.
Commands
Core (local-only, no credentials)
| Command | Description |
|---|---|
init [--ide \x3Cname>] [--profile \x3Cbalanced|low-friction>] |
Generate starter config |
receipt tail [--n \x3Ccount>] |
Print last N receipts |
watch [--config \x3Cpath>] |
Watch event queue, nudge IDE session on new events |
serve [--config \x3Cpath>] |
Start webhook relay server (localhost only) |
memory list|get|set|search |
Manage agent memory (local file backend) |
keepalive start|stop|status |
Prevent macOS sleep for remote sessions |
Advanced (requires credentials or explicit config)
| Command | Requires | Description |
|---|---|---|
sessions send --agent \x3Cid> --message \x3Ctext> |
openclaw.token |
Send message to agent via gateway |
sessions spawn --task \x3Ctext> |
openclaw.token |
Spawn a new agent session |
sessions list|history|status |
openclaw.token |
Query sessions |
gateway trigger|health|agents |
openclaw.token |
Gateway operations |
poll --rooms \x3Cr> --api-key \x3Ck> --handle \x3Ch> |
Ant Farm API key | Poll rooms for messages |
emit --to \x3Curl> --json \x3Cfile> |
None (user specifies target) | POST event JSON to a URL |
hooks create --webhook-url \x3Curl> |
None (user specifies target) | Create webhook forwarder |
tmux run --cmd \x3Ccommand> |
Allowlisted commands only | Run command in tmux, capture receipt |
exec request|resolve|list |
Allowlisted commands only | Execution approval workflow |
cron add|list|remove|run|status |
openclaw.token |
Scheduled task management |
Configuration
Generated by ide-agent-kit init. All credential fields default to empty.
| Field | Purpose | Default |
|---|---|---|
listen.host |
Webhook server bind address | 127.0.0.1 |
listen.port |
Webhook server port | 8787 |
tmux.allow |
Allowlisted shell commands | [npm test, npm run build, pytest, git status, git diff] |
openclaw.token |
Gateway auth (advanced commands) | empty |
github.webhook_secret |
Verify GitHub webhooks | empty |
Data Access
| Path | Access | Purpose |
|---|---|---|
ide-agent-receipts.jsonl |
append | Audit log of all agent actions |
ide-agent-queue.jsonl |
read/write | Event queue |
ide-agent-kit.json |
read | Runtime configuration (may contain secrets) |
memory/ |
read/write | Local agent memory files |
Source & Verification
- npm: https://www.npmjs.com/package/ide-agent-kit
- Source: https://github.com/ThinkOffApp/ide-agent-kit
- Maintainer: petruspennanen (npm), ThinkOffApp (GitHub)
- License: AGPL-3.0-only
The npm package contains no install scripts (preinstall/postinstall). All code is plain ESM JavaScript. Verify with npm pack --dry-run before installing.
Usage Guidance
This skill is internally consistent with its description, but before installing: (1) verify the npm package owner and inspect the package source (github link is in SKILL.md) or install in an isolated environment/container; (2) review the generated config (ide-agent-kit.json) and tighten the tmux/exec allowlist and webhook settings before enabling network features; (3) only populate openclaw.token, Ant Farm API keys, or webhook secrets when you trust the environment and service endpoints; and (4) prefer not to install global CLIs from unknown publishers on production hosts.
Capability Analysis
Type: OpenClaw Skill
Name: ide-agent-kit
Version: 0.4.1
The skill is classified as suspicious due to its capabilities that, if exploited via prompt injection, could lead to data exfiltration. Specifically, the `emit` and `hooks create` commands allow sending arbitrary data to user-specified external URLs. The skill also reads `ide-agent-kit.json`, which the documentation explicitly states 'may contain secrets'. A malicious prompt could instruct the agent to read this configuration file and exfiltrate its contents using the `emit` command. While shell command execution (`tmux run`, `exec`) is present, it is heavily mitigated by a strict default allowlist and an approval flow, reducing its immediate malicious potential. The `SKILL.md` is transparent about these capabilities, but the combination of reading potentially sensitive local files and the ability to send data to arbitrary external endpoints constitutes a significant risk.
Capability Assessment
Purpose & Capability
Name/description match the requested artifacts: the skill is an npm-distributed Node CLI that implements a local filesystem message bus and optional webhook/gateway features. Requiring the node binary and offering an npm install is coherent with the stated purpose. Optional tokens (openclaw.token, Ant Farm API key, GitHub webhook secret) are only referenced for the advanced features they pertain to.
Instruction Scope
SKILL.md instructions focus on running the CLI, using local queue files (./ide-agent-queue.jsonl, receipts), and configuring optional network features. It documents which commands make network calls and which remain local. The only filesystem access described is the working-directory queue and generated config (ide-agent-kit.json), which is consistent with the tool's function.
Install Mechanism
Installation is via the npm package 'ide-agent-kit' (creates ide-agent-kit binary). Using a public npm package is expected for a Node CLI but carries the usual supply-chain risk—package contents are written to disk and executed. No external arbitrary download URLs or extract-from-URL behavior are present, but verify package provenance before installing globally.
Credentials
The skill declares no required environment variables and the SKILL.md only references credentials/config fields that are optional and scoped to advanced features (openclaw.token, Ant Farm API key, github.webhook_secret). This is proportionate to the described functionality.
Persistence & Privilege
No elevated persistence is requested (always: false). The skill does not claim to modify other skills or system-wide agent settings. It writes its own config and queue files in the working directory, which is expected behavior for a local-first CLI.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install ide-agent-kit - After installation, invoke the skill by name or use
/ide-agent-kit - Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.4.1
Fix display name
v0.4.0
ACP session orchestration, message caps, review fixes
v0.3.2
- Updated required environment variables in metadata; now lists none.
- Removed the "primaryEnv" field from metadata.
- No changes to functionality or user-facing documentation.
v0.3.1
v0.3.1 - Room polling, message rate limits, ack-only filtering, file-based notifications
- Clarified and expanded the skill description, highlighting use cases such as event sharing, webhook ingestion, room polling, cron scheduling, and tmux command execution.
- Added required environment variables (OPENCLAW_TOKEN, GITHUB_WEBHOOK_SECRET, ANTFARM_API_KEY) for advanced features.
- Updated metadata to improve visibility of primary environment variable and homepage.
- No file or code changes detected; changelog reflects documentation and metadata improvements only.
v0.2.2
No changes detected in this version.
- Version bumped from 0.2.0 to 0.2.2 with no file changes.
- No new features, fixes, or updates included.
v0.2.1
- Updated skill metadata to add homepage field and remove unused environment variable requirements.
- No changes to functionality or behavior; version number increment only.
- Documentation and configuration instructions remain the same.
v0.2.0
- Added new configuration file: clawhub.json
- Removed old configuration file: metadata.json
- No changes to the main skill documentation or features
- Internal config structure updated for improved management
v1.0.2
- Added metadata.json file for skill metadata.
- No changes to core functionality or documentation.
v1.0.1
- Added detailed security model and explicit command/network behavior documentation to the SKILL.md.
- Declared core/advanced command separation; clarified that advanced features require explicit credentials and opt-in.
- Added environment variable documentation for authentication credentials (`OPENCLAW_TOKEN`, `GITHUB_WEBHOOK_SECRET`, `ANTFARM_API_KEY`).
- Expanded configuration guidance, stating all credential fields are empty by default and no network connections occur unless configured.
- Updated and refined tables for commands, configuration fields, and data access paths.
- Included "Source & Verification" section with links and verification notes.
v1.0.0
Initial release of ide-agent-kit.
Lightweight coordination layer for agentic coding workflows. Gives agents a workspace-local message bus for handoffs and
a receipts-first log for traceability. Zero dependencies.
Four connectivity modes (all composable):
- Local filesystem bus (default) — inter-agent messaging on the same machine, no network
- Webhook relay — accept GitHub/GitLab webhooks as local events
- Ant Farm room polling — cross-machine coordination via room messages
- GitHub events — repo activity translated into local receipts
Defaults are safe: local only, no external connections unless you enable them.
Quickstart:
npm install -g ide-agent-kit
ide-agent-kit init --ide claude-code
ide-agent-kit serve
ide-agent-kit receipt tail
Links:
- npm: https://www.npmjs.com/package/ide-agent-kit
- Source: https://github.com/ThinkOffApp/ide-agent-kit
- SKILL.md: https://github.com/ThinkOffApp/ide-agent-kit/blob/feat/openclaw-skill/SKILL.md
Metadata
Frequently Asked Questions
What is IDE Agent Kit?
Filesystem message bus and webhook relay for multi-agent IDE coordination. Use when agents need to share events, poll Ant Farm rooms, receive GitHub/GitLab w... It is an AI Agent Skill for Claude Code / OpenClaw, with 583 downloads so far.
How do I install IDE Agent Kit?
Run "/install ide-agent-kit" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is IDE Agent Kit free?
Yes, IDE Agent Kit is completely free (open-source). You can download, install and use it at no cost.
Which platforms does IDE Agent Kit support?
IDE Agent Kit is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created IDE Agent Kit?
It is built and maintained by Petrus Pennanen (@thinkoffapp); the current version is v0.4.1.
More Skills