Runtime AI governance: decision gates, human approval, hash-chained audit, compliance mapping.
Enterprise AI governance through the Model Context Protocol.
GIA is a production governance engine that gives AI agents enforceable decision controls, compliance scoring, immutable audit chains, and human-in-the-loop gates. Built for organizations operating under NIST, FedRAMP, CMMC, EU AI Act, and SOC 2 requirements.
29 MCP tools. One integration point. Works with Claude Desktop, Claude Code, OpenAI Agent Builder, and any MCP-compatible client.
npx gia-mcp-server
Or install globally:
npm install -g gia-mcp-server
gia-mcp-server
The server connects to the hosted GIA engine at https://gia.aceadvising.com. Configure your API key:
GIA_API_KEY=your-key npx gia-mcp-server
Add to your claude_desktop_config.json:
{
"mcpServers": {
"gia-governance": {
"command": "npx",
"args": ["-y", "gia-mcp-server"],
"env": {
"GIA_API_KEY": "your-key"
}
}
}
}
claude mcp add gia-governance -- npx -y gia-mcp-server
Point to the Streamable HTTP endpoint:
https://gia.aceadvising.com/mcp
npx -y @smithery/cli install @knowledgepa3/gia-mcp-server --client claude
| Tool | Description |
|---|---|
classify_decision | Classify agent decisions as Mandatory, Advisory, or Informational |
approve_gate | Human-in-the-loop approval for Mandatory gates |
evaluate_threshold | Compute escalation health (Storey Threshold) |
score_governance | Weighted governance scoring (Integrity, Accuracy, Compliance) |
| Tool | Description |
|---|---|
audit_pipeline | Query the hash-chained forensic audit ledger |
verify_ledger | Verify SHA-256 chain integrity from genesis |
map_compliance | Map controls to NIST AI RMF, EU AI Act, ISO 42001, NIST 800-53 |
assess_risk_tier | EU AI Act risk tier classification |
generate_report | Governance status reports (summary, detailed, executive) |
| Tool | Description |
|---|---|
seal_memory_pack | Create immutable, TTL-bound knowledge artifacts |
load_memory_pack | Load packs with trust level and role enforcement |
transfer_memory_pack | Governed knowledge transfer between agents |
compose_memory_packs | Merge packs with risk contamination rules |
distill_memory_pack | Extract governance patterns from usage history |
promote_memory_pack | Promote packs to higher trust levels after review |
| Tool | Description |
|---|---|
monitor_agents | Agent health, repair history, failure counts |
srt_run_watchdog | Infrastructure health probes (API, disk, memory, TLS, DB, DNS) |
srt_diagnose | Incident diagnosis with playbook matching |
srt_approve_repair | Human-approved repair execution |
srt_generate_postmortem | Structured incident postmortems with TTD/TTR metrics |
| Tool | Description |
|---|---|
gia_scan_environment | Scout swarm for environment detection |
gia_list_packs | List remediation, patrol, hardening, and audit packs |
gia_dry_run_pack | Preview pack execution with blast radius analysis |
gia_apply_pack | Execute remediation with mandatory human approval |
gia_run_patrol | Read-only posture checks and compliance audits |
| Tool | Description |
|---|---|
record_value_metric | Track time saved, risks blocked, autonomy levels |
record_governance_event | Log gates, drift prevention, violations blocked |
generate_impact_report | Economic + governance ROI reporting |
system_status | Engine health, uptime, configuration |
GIA enforces governance through three layers:
Every governance action is recorded in a SHA-256 hash-chained audit ledger that can be independently verified.
Built by Advanced Consulting Experts (ACE) — a Service-Disabled Veteran-Owned Small Business (SDVOSB).
GIA was designed by William J. Storey III, a 17-year Information System Security Officer with experience across DoD contracts and U.S. Army Ranger Battalion operations. The same discipline applied to securing classified systems now governs AI agent workforces.
MIT
Source-derived launch command. Check the maintainer’s required arguments and credentials before running:
npx -y gia-mcp-serverMerge this template into ~/Library/Application Support/Claude/claude_desktop_config.json. Keep existing servers. Add any arguments, credentials, and permissions required by the maintainer; this template has not been install-tested.
{
"mcpServers": {
"io-github-knowledgepa3-gia-mcp-server": {
"command": "npx",
"args": [
"-y",
"gia-mcp-server"
]
}
}
}Restart Claude Desktop completely for changes to take effect. Confirm the server appears connected in the client’s tool list, then try a read-only example from its documentation.
Claude Desktop setup referencegia-mcp-servernpmGIA — Governed Intelligence Architecture works with any MCP-compatible client. Copy the config snippet from the Configuration section above and add it to the file shown for your client, then restart the application.
~/Library/Application Support/Claude/claude_desktop_config.jsonRestart Claude Desktop completely for changes to take effect.~/.cursor/mcp.jsonRestart Cursor for changes to take effect..vscode/mcp.jsonReload VS Code window for changes to take effect.~/.codeium/windsurf/mcp_config.jsonRestart Windsurf for changes to take effect..mcp.jsonSave at the project root, then start Claude Code in that project and review the MCP server approval prompt. Keep real credentials out of shared files.