raffkin
v1.0.0Raffkin — agentic SOC analyst skill suite. Investigates and triages Exabeam New-Scale alerts and cases through the Exabeam MCP: evidence gathering, entity pivots, threat/false-positive verdict, and action — with containment recommended for human approval and dismiss/close gated by the host agent's tool-approval rules.
By ExabeamLicense: Apache-2.06 GitHub starsUpdated last week
Directory evidence
- Runtimes
- Codex and Claude Code
- Parsed components
- 4 skill or MCP entries
- Source updated
- Oct 13, 2026
- Manifest status
- Canonical path parsed
The directory validates manifest shape and source location. It does not execute the plugin or provide a security endorsement. Review the indexing methodology →
Install raffkin for Codex and Claude Code
codex plugin marketplace add IchenDEV/agent-plugin-mkt
codex plugin marketplace upgrade agent-plugin-marketplace
codex plugin add raffkin@agent-plugin-marketplacePaste and run these commands in a terminal with Codex. They add and refresh the PluginsMP catalog, then install this plugin.
The installer fetches third-party code from the source repository shown on this page. This directory validates manifest structure and source location, but does not perform a security audit; review the manifest, components, and source before installing.
Get the source manually
git clone https://github.com/open-agent-ai-security/raffkinClone the source repository, then follow its setup instructions to add the plugin to a compatible client. The plugin root is plugin/.
Plugin files
├── .codex-plugin/plugin.json├── .claude-plugin/plugin.json├── skills/rule-tuning/SKILL.md├── skills/soc-investigate/SKILL.md├── skills/triage-cases/SKILL.md└── .mcp.json
Included Skills3
Find NOISY detection rules in Exabeam New-Scale — noisy, not merely loud — and propose specific, actionable tuning. Use when the analyst asks to "find noisy rules", "what's generating false positives", "tune detections", "reduce alert noise", "why is this rule so loud", or "which rules waste our time". Reads the rule inventory and case/detection history through the Exabeam MCP, separates high-volume-low-precision rules from high-volume-high-precision ones, and proposes tuning mapped to real Exabeam mechanics (context table, exclusion rule, or the rule's own filter/scope/maturity settings). Read-only and propose-only: the rule-write tools the MCP exposes are denied on both hosts — detection engineering applies the change. Requires the Exabeam MCP server to be configured.
Investigate and triage a security alert or case in Exabeam New-Scale, end to end. Use when the analyst hands over an alert/case ID or payload, or asks to "investigate", "triage this alert/case", "work this alert", or "is this a real threat?" — for one item at depth, not a whole queue (sweeping the queue is the `triage-cases` skill). Gathers evidence through the Exabeam MCP, correlates activity into a timeline, maps to MITRE ATT&CK, reaches a threat / false-positive verdict, takes the non-destructive terminal action (open a case, dismiss an alert, write case notes), and recommends any containment for the analyst to approve. Requires the Exabeam MCP server to be configured.
Prioritize a queue of open Exabeam New-Scale cases — decide what needs a human's attention first. Use when the analyst asks to "triage the queue", "what should I look at", "what needs attention", "sweep the open cases", "morning triage", or "prioritize", rather than handing over one specific case ID. Reads the open case queue through the Exabeam MCP, clusters cases by attack-shape, ranks them by corroborated signal (with the risk score as one tunable input, not the sole one), and returns a short "start here" list plus the noise clusters worth tuning. Read-only during the sweep: it prioritizes and flags, and may call an obvious verdict, but never auto-writes across the queue. Hand a single case to soc-investigate; hand a noise cluster to rule-tuning. Requires the Exabeam MCP server to be configured.
MCP servers1
- command
- uv
- args
- run --quiet --locked connector/exabeam-mcp-bridge.py
- cwd
- .
Plugin manifests2
{
"name": "raffkin",
"version": "1.0.0",
"description": "Raffkin — agentic SOC analyst skill suite. Investigates and triages Exabeam New-Scale alerts and cases through the Exabeam MCP: evidence gathering, entity pivots, threat/false-positive verdict, and action — with containment recommended for human approval and dismiss/close gated by the host agent's tool-approval rules.",
"author": {
"name": "Exabeam",
"url": "https://www.exabeam.com/"
},
"homepage": "https://github.com/open-agent-ai-security/raffkin",
"repository": "https://github.com/open-agent-ai-security/raffkin",
"license": "Apache-2.0",
"keywords": [
"soc",
"soc-analyst",
"alert-triage",
"exabeam",
"ai-agent-security",
"mcp",
"codex"
],
"skills": "./skills/",
"mcpServers": "./.mcp.codex.json",
"interface": {
"displayName": "Raffkin — Agentic SOC Analyst",
"shortDescription": "Triage Exabeam New-Scale alerts and cases end to end — evidence, verdict, and a guarded write path.",
"category": "Security"
}
}{
"name": "raffkin",
"version": "1.0.0",
"description": "Raffkin — agentic SOC analyst skill suite. Investigates and triages Exabeam New-Scale alerts and cases through the Exabeam MCP: evidence gathering, entity pivots, threat/false-positive verdict, and action — with containment recommended for human approval and dismiss/close gated by the host agent's tool-approval rules.",
"author": {
"name": "Exabeam",
"url": "https://www.exabeam.com/"
},
"homepage": "https://github.com/open-agent-ai-security/raffkin",
"repository": "https://github.com/open-agent-ai-security/raffkin",
"license": "Apache-2.0",
"keywords": [
"soc",
"soc-analyst",
"alert-triage",
"exabeam",
"ai-agent-security",
"mcp",
"claude-code"
],
"skills": "./skills/"
}For maintainers
If you maintain this plugin, link to this source-backed listing from your README so users can review its manifest and indexed components.
[raffkin on Agent Plugins Marketplace](https://pluginsmp.com/plugins/raffkin)