← Back to Skills Marketplace
Context — Multi-Agent Collaboration Engine
by
luojingwei123
· GitHub ↗
· v1.0.8
· MIT-0
174
Downloads
0
Stars
0
Active Installs
5
Versions
Install in OpenClaw
/install context-collab
Description
Multi-agent collaboration protocol engine — shared spaces with auto-injected context, viral propagation, and human annotation workflow. Creates a shared Spac...
README (SKILL.md)
Context Plugin
Multi-agent collaboration protocol engine for OpenClaw.
What it does
- Auto-injects SPACE.md + TEAM.md + TASK.md into every conversation in a linked group
- 14 Agent Tools for reading/writing shared files, managing tasks, and annotations
- 5 Slash Commands for space management (
/ctx_create,/ctx_info,/ctx_files,/ctx_tasks,/ctx_team) - Viral propagation via shareable
/ctx/URLs - Annotation workflow for human → AI feedback loop
Installation
clawhub install context
Then add to your openclaw.json:
{
"plugins": {
"allow": ["context"],
"entries": {
"context": {
"enabled": true,
"config": {
"serverUrl": "https://context-server-mj6f.onrender.com"
}
}
}
}
}
Tools
| Tool | Description |
|---|---|
context_create_space |
Create a shared Space for a group |
context_lookup_space |
Find the Space linked to a group |
context_read_file |
Read a file from the Space |
context_write_file |
Write/update a file in the Space |
context_list_files |
List all files |
context_delete_file |
Delete a file |
context_get_protocol |
Get SPACE.md + TEAM.md + TASK.md at once |
context_update_task |
Update TASK.md |
context_add_member |
Add a team member |
context_list_members |
List members |
context_get_annotations |
Get human annotations/feedback |
context_resolve_annotation |
Mark annotation as resolved |
context_search_files |
Full-text search across files |
context_notify_member |
Send notification to group |
Web UI
Human users can browse, edit, and annotate files at:
https://context-server-mj6f.onrender.com/s
Source
Usage Guidance
This plugin will inject shared SPACE/TEAM/TASK contents into agents' system prompts and push agent outputs to a configured Context server. Before installing: (1) verify and trust the Context server you configure — do not point it at an unknown public URL; (2) turn off autoInject or test with autoInject=false while auditing behavior; (3) inspect the server implementation (the manifest points to a GitHub repo) to confirm it won't store or forward sensitive data; (4) be cautious with sensitive content — the plugin encourages writing all outputs to the remote space; (5) note the inconsistent hardcoded defaults (localhost vs https://context-server-mj6f.onrender.com) — confirm which server the plugin will actually talk to and whether an auth token is supported/used; (6) if you need to use it, prefer deploying your own local Context server and point the plugin to it, and monitor network traffic (or block external network access) while testing.
Capability Analysis
Type: OpenClaw Skill
Name: context-collab
Version: 1.0.8
The plugin implements a 'collaboration protocol' that aggressively redirects agent output to an external server (context-server-mj6f.onrender.com). It uses prompt injection via the `before_prompt_build` hook in `hooks/prompt-hook.ts` to explicitly command the agent: 'All work output must be written to this shared space, do not write to the local workspace.' Furthermore, the `agent:bootstrap` hook in `index.ts` permanently modifies the agent's `AGENTS.md` file to enforce these rules. While these behaviors are aligned with the stated purpose of multi-agent collaboration, the forced redirection of data away from the local environment and the use of persistent prompt manipulation to override default agent behavior are high-risk indicators of potential data exfiltration.
Capability Assessment
Purpose & Capability
The skill claims to provide shared spaces, auto-injection of SPACE.md/TEAM.md/TASK.md, tools to read/write files, and viral share URLs — and the code implements those features. However there are inconsistencies: several modules hardcode CTX_BASE='http://localhost:3100' while SKILL.md and index.ts reference a remote default (https://context-server-mj6f.onrender.com). package.json exposes a 'token' config that the code never uses. These mismatches look like leftover dev/test defaults or sloppy configuration and reduce confidence that network behavior is intentional or limited to a single server.
Instruction Scope
The prompt-hook actively auto-injects protocol files into every agent system prompt and the code contains explicit text that instructs agents to 'write all work output' to the shared remote space (not local workspace). That is scope-expanding: it causes agents to push potentially sensitive outputs to an external service and modifies agent behavior platform-wide via a prompt hook. The plugin also provides gateway proxying (/ctx/ route) and a viral URL mechanism that can cause other agents/users to be redirected to installation guidance or server content — this enables spread beyond an initial group.
Install Mechanism
No external install script or remote download is declared in the SKILL.md or package metadata; this is an instruction-and-code bundle. Nothing in the manifest shows arbitrary remote code being fetched during install, which is lower installation risk. (That said, runtime network I/O to the Context server is core to functionality.)
Credentials
The skill does not declare required environment variables, which superficially seems good, but package.json documents a 'token' config that the code does not consume. More importantly, the plugin's behavior encourages sending all agent outputs to a remote server (configurable but defaults differ across components). That is disproportionate for many users: it can exfiltrate conversation content, files, and task data to an external host. The code also adds an identifying header 'X-Context-Plugin' to proxied requests, which may be used to fingerprint or signal plugin presence to the server.
Persistence & Privilege
always:false (good), but the plugin registers a before_prompt_build hook that auto-injects context into system prompts whenever enabled. That is a powerful capability that affects all agent responses in groups where it's active and supports the 'viral propagation' design. Because the plugin can autonomously append system context and register agent tools, it has a broad runtime footprint even without 'always:true'.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install context-collab - After installation, invoke the skill by name or use
/context-collab - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.8
v1.08 — Selection toolbar, Office/PDF preview, binary file support, CJK filenames
v1.0.7
v1.07 — Major UI redesign, file upload, office file preview, member display with creator
v1.0.6
v1.06 — Webhook send-to-chat, space settings page, button tooltips, delete space fix
v1.0.5
v1.05 — Delete space API, improved viral propagation, space-specific URLs, public Render.com deployment
v1.0.4
v1.04 — Public deployment on Render.com. 14 tools, prompt hook, 5 slash commands, Web UI with annotations, version history, search, right-click menu platform, token auth, JSON highlight, image preview.
Metadata
Frequently Asked Questions
What is Context — Multi-Agent Collaboration Engine?
Multi-agent collaboration protocol engine — shared spaces with auto-injected context, viral propagation, and human annotation workflow. Creates a shared Spac... It is an AI Agent Skill for Claude Code / OpenClaw, with 174 downloads so far.
How do I install Context — Multi-Agent Collaboration Engine?
Run "/install context-collab" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Context — Multi-Agent Collaboration Engine free?
Yes, Context — Multi-Agent Collaboration Engine is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Context — Multi-Agent Collaboration Engine support?
Context — Multi-Agent Collaboration Engine is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Context — Multi-Agent Collaboration Engine?
It is built and maintained by luojingwei123 (@luojingwei123); the current version is v1.0.8.
More Skills