wrg-mcp-server 1.0.4

WRG MCP server - exposes WinstonRedGuard tools to Claude and AI agents

wrg_mcp_server

MCP (Model Context Protocol) server exposing the WinstonRedGuard monorepo to Claude and other MCP-compatible AI agents. Built on FastMCP — registers tools from every active WRG app so an agent can inspect the repo, run pipelines, query memory, and call remote services without shelling out.

Published:

• PyPI: wrg-mcp-server (v1.0.2)
• MCP Registry: io.github.WRG-11/wrg-mcp-server (status: active)

Transports

wrg-mcp-server --transport stdio              # Claude Desktop / Claude Code
wrg-mcp-server --transport streamable-http    # default HTTP (recommended)
wrg-mcp-server --transport sse                # legacy HTTP

Flags: --host 0.0.0.0 · --port 8080 · --mcp-path /mcp

Install

cd apps/wrg_mcp_server
pip install -e .                  # core: MCP + local tools only
pip install -e ".[remote]"        # adds httpx for site_* / pulseboard_* tools
pip install -e ".[threat-intel]"  # adds wrg_threat_intel for attack_surface/ransomware/darkweb tools
pip install -e ".[dev]"           # pytest + pytest-asyncio

Tools exposed

Local (subprocess-backed, always available)

Tool	What it does
connector_status	Report which remote services are configured
app_list, app_info	Query app_registry/data/registry.json
governance_run	Execute governance_check across one or all apps
release_check	Run the tools/release_check.ps1 gate
pipeline_list, pipeline_show, pipeline_run	wrg_pipeline DAG operations
pulse_check	Invoke wrg-pulse check
memory_get, memory_set, memory_list, memory_search	wrg_memory key-value access
research_history, research_report, research_scan, research_watch, research_scan_summary	research_motor runs and artifacts
research_motor_healthz, research_motor_scan_create, research_motor_scan_get	research_motor HTTP API v1 over localhost
vault_audit	wrg_vault audit ledger inspection
scheduler_task_list, scheduler_tick_dry_run	wrg_scheduler inspection

Round 39 — silo-app expansion (6 apps × 2 tools = 12 tools)

AI fingerprint (wrg_ai_fingerprint)

Tool	What it does
ai_fingerprint_scan	Scan a path for AI-generated code signals; supports min_score, exclude[]
ai_fingerprint_detectors	List registered detectors and their weights
ai_fingerprint_sigma_emit	Convert fingerprint scan JSON into Sigma YAML rules (R42 — wrg_ai_fingerprint_sigma emit; pass input_file or scan_json, optional date_override)

DevGuard (wrg_devguard)

Tool	What it does
devguard_scan	Run policy / secrets / crypto scans on a path; empty scan_types runs combined check
devguard_baseline	List configured policy profiles (baseline + strict) and presence

Security suite (wrg_security_suite) — security_suite_run is mutation-gated

Tool	What it does
security_suite_run	Run code / person / network / full scan (mutation — requires WRG_MCP_ALLOW_MUTATIONS=1)
security_suite_report	Read a scan report by scan_id from apps/wrg_security_suite/reports/ (read-only)

Rule lab (rule_lab)

Tool	What it does
rule_lab_test	Simulate a rule set against sample contexts (simulate --json)
rule_lab_list	List rule files under $WRG_RULE_LAB_DIR or <repo>/.wrg/rules

Data janitor (data_janitor) — data_janitor_sweep mutation-gated when dry_run=False

Tool	What it does
data_janitor_sweep	Scan (dry-run) or clean build artifacts; non-dry requires WRG_MCP_ALLOW_MUTATIONS=1
data_janitor_orphans	Preview orphan / build-artifact targets (read-only)

Notifier (wrg_notifier3) — notifier_send is mutation-gated

Tool	What it does
notifier_send	Dispatch a message to a configured channel (mutation — requires WRG_MCP_ALLOW_MUTATIONS=1)
notifier_channels	Introspect available channel adapters (read-only)

Round 50 — INFO_OPS extension

Tool	What it does
info_ops_detect	Query INFO_OPS actor corpus (default modus=info-ops); enrich each match with linked incidents + Sigma rules; reverse-lookup via mitre_technique filter

info_ops_detect usage

Default — list all INFO_OPS actors with Sigma + incidents:

info_ops_detect()
# -> 3 actors (russia_nexus_info_ops, china_nexus_spamouflage, iran_nexus_apt35),
#    ~21 Sigma rules total, linked incidents per actor

Specific actor lookup:

info_ops_detect(actor_id="russia_nexus_info_ops")
# -> single-actor detection state (Sigma rules + 2 incidents)

Reverse-lookup — "which info-ops actors use T1656 Impersonation?":

info_ops_detect(mitre_technique="T1656")
# -> china_nexus_spamouflage + iran_nexus_apt35 (R52);
#    russia_nexus_info_ops excluded (uses T1078 Valid Accounts)

Arastirma Ussu (opt-in via env)

Tool	What it does
arastirma_ask	Ask a question to the Arastirma Ussu knowledge base
arastirma_doc_search	Search documents in the knowledge base
arastirma_web_search	Web search through Arastirma Ussu
arastirma_memory_search	Search memory entries

Trading (opt-in via env)

Tool	What it does
trading_analyze	Full multi-agent trading analysis for a ticker
trading_quick_signal	Fast RSI/price signal for a ticker

OSINT

Tool	What it does
maigret_search	Username search across 3000+ sites (Maigret)

Threat-intel (opt-in via [threat-intel] extra)

Tool	What it does
attack_surface_passive	Passive attack surface reconnaissance
ransomware_lookup	Ransomware group/victim lookup
darkweb_brand_watch	Dark web brand mention monitoring

Ransom-radar

Tool	What it does
ransom_radar_tick	Run a ransom-radar feed tick
ransom_radar_status	Check ransom-radar watchlist status

Other

Tool	What it does
polymarket_event_signal	Polymarket event signal analysis
research_deep	Deep research with AI research platform

Remote (HTTP, opt-in via env)

Tool	Upstream
site_health, site_get, site_post	Company site API (WRG_SITE_BASE_URL)
pulseboard_health, pulseboard_list_repos, pulseboard_add_repo, pulseboard_delete_repo, pulseboard_get_pulse	pulseboard dashboard (WRG_PULSEBOARD_BASE_URL)

Remote tools return {"ok": false, "error": "httpx not installed — remote tools unavailable"} when the [remote] extra is not installed.

research_motor HTTP API tools

Start the research_motor API separately, then point the MCP server at it:

cd apps/research_motor
pip install -e ".[api]"
set RESEARCH_MOTOR_API_KEY=replace-me
research-motor serve --host 127.0.0.1 --port 8080

Configure the MCP server environment:

set WRG_RM_API_BASE_URL=http://127.0.0.1:8080
set WRG_RM_API_KEY=replace-me

Example MCP tool payloads:

{"target": "example.com", "mode": "domain"}

{"scan_id": "sample-scan-abc123"}

Environment

Repo discovery

Variable	Default	Purpose
WRG_REPO_ROOT	auto-detect (walk up until apps/ + CLAUDE.md)	Required when installed from wheel outside the monorepo

Mutation gate (default: off)

State-changing tools (memory_set, pipeline_run, security_suite_run, data_janitor_sweep non-dry, notifier_send) refuse to execute unless:

WRG_MCP_ALLOW_MUTATIONS=1

This prevents an MCP client from silently writing memory or launching pipelines on a read-only connection.

Remote service config

Per service (SITE / PULSEBOARD), prefix with WRG_<SERVICE>_:

Variable	Default	Purpose
*_BASE_URL	—	Enables the service (unset = service disabled)
*_TOKEN	—	Bearer token for Authorization header
*_AUTH_HEADER	Authorization	Override header name
*_AUTH_SCHEME	Bearer	Override token scheme
*_SESSION_COOKIE	—	Optional Cookie header
*_EXTRA_HEADERS	—	JSON object of extra headers
*_TIMEOUT_SECONDS	WRG_HTTP_TIMEOUT_SECONDS (20.0)	Per-request timeout
*_VERIFY_TLS	WRG_HTTP_VERIFY_TLS (true)	TLS verification

Claude Code / Claude Desktop integration

Add to your MCP client config:

{
  "mcpServers": {
    "wrg": {
      "command": "wrg-mcp-server",
      "args": ["--transport", "stdio"],
      "env": {
        "WRG_REPO_ROOT": "D:\\dev\\WinstonRedGuard",
        "WRG_MCP_ALLOW_MUTATIONS": "0"
      }
    }
  }
}

Architecture

FastMCP server
├── server.py            — tool registration, remote HTTP dispatch
├── config.py            — ServiceConfig / AppConfig from env (frozen dataclasses)
├── http_utils.py        — URL builder, response parser
├── local_tools.py       — subprocess wrappers for WRG CLIs (~20 tools)
└── cli.py               — argparse entry point

Local tools use subprocess.run with stdin=DEVNULL (not asyncio subprocess) — avoids a Windows pipe-blocking deadlock under anyio. Tool dispatch is wrapped in anyio.to_thread.run_sync so the MCP event loop stays responsive.

Tests

pytest -q

Status

Production — 1045 lines, covers every active WRG app, drives the mcp__wrg__* tools visible in connected Claude sessions.