Skill profile Review candidate Next best action: Check evidence, trust, and trial before install

~/agentstaffer / skills / github-nduckmink-arkon

nduckmink/arkon

Arkon: Enterprise AI Knowledge Hub & MCP Server. Self-hosted knowledge base for teams to manage RAG contexts, access policies, and AI skills. Connect Claude and other LLMs via Model Context Protocol (MCP) for automated, secure organizational knowledge integration.

githubskillMCP & Tool IntegrationMCPMCP clientsClaude Desktop

Fit & Capabilities

What this candidate is suited for, how the fit score breaks down, and which protocols or skills are visible.

51 / 100 fit

Fit score51 / 100Capability fit: 79/100

Best for1 usesTeams standardizing agent tool access through MCP

Protocols1 visibleMCP

Capabilities3 listedRAG / MCP integration

Fit detailOpen the full fit breakdown.

Capability fit79 / 100

MCP & Tool Integration

Trust50 / 100

Needs review

Integration readiness25 / 100

MCP ready

Maintenance signal53 / 100

908 stars

Community proof50 / 100

Community review pending

Trial readiness29 / 100

Run an MCP handshake trial

Best for1 uses

Teams standardizing agent tool access through MCP

Protocols1

MCP

Capabilities3

RAGMCP integrationReusable skill package

Evidence

Source traction, reputation reviews, verified resources, and missing signals grouped once for the profile decision.

3 quality signals

Source traction908 starsStars

Reputation & reviewsPending0 community / 0 trials

Verified resources3 / 3 presentMCP config available

Evidence gaps3 openOpen evidence detail before promotion

Evidence detailOpen the full evidence ledger.

Source traction

Stars: 908
Forks: 210
Last activity: 2026-05-28

Reputation & reviews

External reputation: Pending; source traction tracked separately
Community reviews: 0
Trial runs: 0
Traction evidence: See source traction card

Verified resources

Source URL: Present
Install manifest: Present
MCP config: Present

Evidence gaps

External reputation: Pending; source traction tracked separately
Sandbox trial: Not recorded
Community review: Not recorded

Risk & Safety

Risk signals, permission requirements, and required checks before install, trial, hire, or production use.

Trial review

Risk signals

No community evaluationmedium

AgentStaffer users have not submitted a review yet.

Third-party reputation pendingmedium

Third-party review or discussion signal is pending. Source stars, downloads, likes, and activity are tracked separately as source traction.

Marketplace review notereview

matches:mcp,skill,skills,rag

Marketplace review notereview

protocol:mcp

Marketplace review notereview

capabilities:RAG,MCP integration,Reusable skill package

Safety requirements

Sandbox workspacerequired

Run the first trial in a throwaway directory or isolated environment.

MCP-capable clientrequired

Use Claude Desktop, Cursor, VS Code, or another client that can load MCP server configs.

Quality evidence

Protocol metadatapositive

MCP support detected.

License visiblepositive

License metadata: Other.

Recent source activitypositive

Updated about 141 days ago.

Required checks

Verify install sourcerequired

Open install.json and source docs before running any command.

Collect user evaluationreview

No AgentStaffer community review is attached yet.

Run sandbox trialrequired

Create a narrow, observable trial before install, hire, or promotion.

Security verdict

Install path is visible, but review permissions and run a sandbox task first.

Open trust review

Safe action rule

This is an evidence-based AgentStaffer review summary, not a malware guarantee. Always inspect source, install manifest, permissions, and trial results before local execution.

Installation

One install command/config, with client-specific notes kept short.

MCP ready

Command

Review this AgentStaffer install manifest before installing: https://agentstaffer.com/skills/github-nduckmink-arkon/install.json

Setup guide

MCP config

{
  "mcpServers": {
    "nduckmink-arkon": {
      "command": "REPLACE_WITH_AGENT_COMMAND",
      "args": [],
      "env": {
        "AGENTSTAFFER_PROFILE": "https://agentstaffer.com/skills/github-nduckmink-arkon"
      },
      "_agentstaffer": {
        "note": "AgentStaffer could detect MCP support, but could not infer a runnable command. Check the source documentation before use.",
        "sourceUrl": "https://github.com/nduckmink/arkon"
      }
    }
  }
}

MCP config

Client notes

Claude Desktop

Copy the MCP config from this page or download the MCP config file.

Cursor / VS Code

Add the downloaded MCP config to your editor MCP settings.

Dify / n8n / Langflow

Download the AgentStaffer install file and attach it to your workflow notes.

Local / Terminal

Use the source repository instructions.

Safety

Run in a sandbox or throwaway workspace.
Use non-production data and temporary credentials.
Capture output, logs, install friction, and any permission requests.

Trial Plan

Status: Not started. Record trial evidence before production promotion.

not started

Run an MCP handshake trial

Install the MCP config, list available tools, and call one read-only tool with mock data.

Expected output

A successful client connection plus a read-only tool result and clear permission notes.

Success criteria

Config loads cleanly
Tools are named clearly
No secret leakage
Read-only test succeeds

Required review

Sandbox workspace: Run the first trial in a throwaway directory or isolated environment.
MCP-capable client: Use Claude Desktop, Cursor, VS Code, or another client that can load MCP server configs.

Promotion rule

Blocked until trial evidence exists

Open trial workspace Trial JSON

Lifecycle & Alternatives

Human-readable review history and nearby alternatives stay compact here; full events and graph live in the complete review materials.

0 records

Reviews & HistoryReview history

Indexed for hiring review

Candidate entered AgentStaffer as a public marketplace listing.

2026-07-14

Current lifecycle stage

Run Trial: Start with a sandbox trial before shortlisting or hiring.

2026-07-14

Latest sandbox trial

No sandbox trial has been recorded, so hire/install decisions should stay conservative.

Pending

Latest community review

No community review has been attached yet.

Pending

Similar CandidatesAlternatives

IvanMurzak/Unity-MCP

AI Skills, MCP Tools, and CLI for Unity Engine. Full AI develop and test loop. Use cli for quick setup. Efficient token usage, advanced tools. Any C# method may be turned into a tool by a single line. Works with Claude Code, Gemini, Copilot, Cursor and any other absolutely for free.

77/100 fit

wshobson/agents

Multi-harness agentic plugin marketplace for Claude Code, Codex CLI, Cursor, OpenCode, and Gemini CLI

75/100 fit

DemonDamon/AgenticX

AgenticX is a unified, production-ready multi-agent platform — Python SDK + CLI (agx) + Studio server + Machi desktop app. Features Meta-Agent orchestration, 15+ LLM providers, MCP Hub, hierarchical memory, avatar & group chat, skill ecosystem, safety sandbox, and IM gateway (Feishu/WeChat).

73/100 fit

NevaMind-AI/memU

Memory for 24/7 proactive agents like OpenClaw.

68/100 fit

ruvnet/ruflo

🌊 The leading agent orchestration platform for Claude. Deploy intelligent multi-agent swarms, coordinate autonomous workflows, and build conversational AI systems. Features enterprise-grade architecture, self-learning swarm intelligence, RAG integration, and native Claude Code / Codex Integration

65/100 fit

xurl skill by openclaw/openclaw

This skill helps you interact securely with the X API via xurl, enabling posts, reads, DMs, searches, and multimedia actions.

79/100 fit

Decision files

Review materials

Scan the decision path first; expand only when the recommendation needs source files, history, alternatives, or dossier detail.

Open profile JSON

Decision path

Profile JSONStartEvidence JSONSignalsTrust JSONRiskMaterialsAll files

01Decision basis

Start with the profile record, then verify score, evidence, and trust before shortlisting.

Profile, score, evidence, trust

02Trial readiness

Use the sandbox plan and setup guide before copying commands or approving local setup.

Trial plan and setup

03Governance trail

Check lifecycle history, alternatives, and the dossier before hire, promotion, or credentials.

History, alternatives, dossier

Full file listOpen ordered resources when the decision needs more evidence.

9 required records stay in review order; companion links are grouped for context.

Required recordsRead these before install, hire, promotion, or credential access.

Profile recordIdentity and resource contract Score detailFit and recommendation score Evidence ledgerSignals, sources, and gaps Trust reviewPermissions and safety Trial guideSandbox task and criteria Setup guideInstall method and client notes Review historyPast reviews and changes AlternativesNearby candidates Readable dossierMarkdown decision brief

Companion links8 supporting records for context and discovery.

Short summaryShort summary MCP configClient configuration ComparisonSide-by-side context Visibility reportVendor claim and evidence gaps claim.jsonCompanion link Resource mapSite resource map Site contextShared review context SitemapDiscovery index

Review order

01Decision recordIdentity, summary, and required review materials
02Score detailFit, recommendation, and decision scoring
03Evidence ledgerSource, reputation, and review gaps
04Full file listEvidence, setup, trust, trial, and history files

Use the compact page for quick review; open the full file list before recommending hire, install, promotion, or credential access.

Vendor path

Claim, verify, report, then feature

Vendor ownership and paid visibility options stay below the review package; evidence scores stay independent.

Start claim

Next stepClaim owner path

Submit public maintainer proof so corrections and paid visibility can be reviewed.

ReportEvidence gaps first

Open vendor visibility report

GuardrailExposure only

Paid placement can influence eligible exposure surfaces only after verification; trust, fit, safety, trial, and promotion scores remain evidence based.

done Indexed

The profile is discoverable with public profile, score, evidence, trust, install, and trial resources.

Open profile

current Claim owner path

Submit public maintainer proof so corrections and paid visibility can be reviewed.

Start claim

locked Verify evidence

Ownership, install metadata, safety posture, trial evidence, and public proof are checked independently.

Open visibility report

current Read visibility report

Use the report to see crawlable surfaces, comparison presence, evidence gaps, and the next fix.

View report

locked Apply for featured visibility

Featured exposure is available only after claim or verification readiness, and never changes scores.

Complete claim first

Commercial products and availabilityOpen claim, verification, report, and featured visibility options.

Claim profilefree

available · Start claim

Start claim

Verification reviewoptional paid

locked · Claim first

Claim first

Visibility reportoptional paid

available · Open report

Open report

Featured visibilitypaid

locked · Complete verification

Complete verification

Review route: OpenAI review tools can start with openai.json; otherwise use site context and agent inventory before opening the finalist's complete review materials before install, hire, or promotion.