Shared Context Mcp — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited Shared Context Mcp (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Install this skill to use the Agent Bridge MCP from any Claude Code session.
When you have multiple AI agent windows open (Claude Code, OpenCode, Codex, etc.), this skill helps you:
hackathon, my-project, review)The mcp-agent-bridge server must be running:
git clone https://github.com/YOUR_USERNAME/mcp-agent-bridge
cd mcp-agent-bridge
npm install && npm startAdd to ~/.claude.json:
{
"mcpServers": {
"agent-bridge": {
"type": "http",
"url": "http://localhost:3721/mcp"
}
}
}identify("claude", session_label="<what you're working on>")list_contexts() # see what's available
join_context("my-project") # join an existing one
create_context("new-feature", description="...") # or create newset_memory("current_task", "implementing OAuth")
set_memory("decisions", '{"db": "postgres", "orm": "drizzle"}', tags=["architecture"])context_status() # peers online + context summary
read_messages() # messages from other agents
get_events(limit=20) # recent activity in this contextcreate_task("Write unit tests", "Cover auth module", assigned_to="opencode")
list_tasks(filter="mine")
update_task("<id>", status="done", notes="All 42 tests passing")save_artifact("api-design", "<content>", type="markdown", description="REST API spec")
get_artifact("api-design")| What you want | Tool |
|---|---|
| Register this session | identify(name, label?, context?) |
| See all workspaces | list_contexts() |
| Switch workspace | join_context(name) |
| Create workspace | create_context(name, description?) |
| Write shared data | set_memory(key, value) |
| Read shared data | get_memory() |
| Message an agent | send_message(to, content) |
| Check messages | read_messages() |
| Create a task | create_task(title, desc, assigned_to) |
| Share code | save_artifact(name, content, type) |
| See recent events | get_events(since?) |
| Who's online | agent_status() |
| Leave workspace | leave_context() |
Any agent write (memory, message, task, artifact) is instantly pushed to all other agents in the same context — no polling needed. Subscribe directly via WebSocket for dashboards or scripts:
node -e "
const {WebSocket}=require('ws');
const ws=new WebSocket('ws://localhost:3721/ws');
ws.on('message',d=>{const m=JSON.parse(d);if(m.type!=='snapshot')console.log(m.context,m.type,m.data)});
"~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.