HARP Project

Evidence-first AI reliability, not autonomous remediation

HARP — Homelab & Hybrid AI Reliability Platform

A human-in-the-loop reliability platform that turns alerts, runbooks, RCAs, session logs, and monitoring context into cited evidence packs. HARP, Nero-Camp, OpenRAG, read-only MCP, LiteLLM, and n8n work together so AI tools can explain incidents while humans keep approval authority.

BM25 + hybrid searchEvidence packs with citationsRead-only MCP agent surfaceOpenRAG semantic retrievalLiteLLM model routingNero-Camp workflow & approval staten8n workflow glueHuman approval before remediationEval harness + golden queries

View Architecture Incident Flow Links & Proof Points

Control-room architecture read-only first

HARPBM25 · hybrid search · evidence packs · MCP

AI clientsCodex, Cursor, Claude Desktop

SignalsPrometheus, Alertmanager, Uptime Kuma

LiteLLMmodel routing gateway

OpenRAGsemantic retrieval

Nero-Campworkflow and approval state

Git + runbooksRCAs, docs, sessions, evals

tests passing

Stage A–B test suite

golden queries

Eval harness corpus

D+H

current stage

OpenRAG hybrid + MCP

0.5.0

MCP version

Read-only tools live

Core Message

The reliability boundary is the product.

HARP is built around a clear split: evidence, workflow state, retrieval, model routing, event glue, and approvals each have one owner.

HARP is evidence

HARP turns Git-tracked runbooks, RCAs, session logs, alerts, metrics, and generated evidence packs into searchable reliability memory with citations and confidence signals.

MCP is read-only

AI tools can ask HARP for context through MCP tools and resources. They do not get shell access, Git writes, kubectl mutation, or remediation power.

Nero owns approval

Nero-Camp owns workflow state, approval, audit timeline, and the human decision record. HARP provides evidence; Nero records the operational decision.

Scope Boundary

What HARP is not.

The deliberate scope is as important as the feature set. These are hard boundaries, not roadmap gaps.

Not autonomous remediation

HARP does not restart services, apply patches, or execute kubectl mutations. Evidence-only means evidence-only.

Not a general chatbot

HARP has no broad operational credentials. Every answer must cite a repo-relative file path or live evidence.

Not a SaaS product

HARP is self-hosted, content-agnostic, and profile-driven. Point it at any Git runbook repo and it indexes that repo.

Not a replacement for monitoring

HARP does not replace Prometheus, Grafana, Alertmanager, or Argo CD. It is a reliability knowledge layer that sits alongside them.

Talk Track

How to explain HARP.

The problem

Incident context is scattered across alerts, logs, runbooks, dashboards, and old RCAs. Copying that manually into AI chats creates weak grounding.

The architecture

Signals enter HARP. HARP searches local BM25 and optional OpenRAG, merges results, and returns cited evidence instead of invented remediation.

The agent interface

MCP standardizes how Codex, Cursor, Claude Desktop, or reviewer agents ask HARP for tools, resources, and prompt templates.

The workflow layer

Nero-Camp consumes HARP evidence, updates task state, keeps approval history, and separates operator decisions from model output.

The enterprise lesson

The homelab shape maps to enterprise platform engineering: retrieval, model routing, agent contracts, workflow orchestration, governance, and evals.

Modules

Explain each part one by one.

HARP

Evidence

Search API, evidence renderer, MCP server, policy labels, eval harness, and OpenRAG connector.

Explain HARP →

OpenRAG

Semantic

Milvus-backed semantic retrieval for symptom-language discovery beyond exact keyword matches.

Explain OpenRAG →

MCP

Agent API

A constrained tool/resource/prompt surface for AI clients that need context, not production power.

Explain MCP →

LiteLLM

Routing

Model gateway for local models or gated cloud providers such as AWS Bedrock when policy allows.

Explain LiteLLM →

Nero-Camp

Approval

Task state, source records, timelines, approval gates, and human closeout decisions.

Explain Nero-Camp →

Authorship

What I built vs. what I integrated.

A clear split between components I authored from scratch and components I selected, configured, and integrated into the architecture.

Component	Role	Evidence
HARP search engine	Built	Pure-Python BM25 + RRF hybrid merge in harp/core/
Evidence pack generator	Built	Alert payload → cited Markdown with action tiers
MCP server	Built	FastMCP 3.x, 5 tools, 4 resources, 3 prompts
Eval harness	Built	29 golden queries, MRR scoring, harp/evals/
RAG export connector	Built	Provider-neutral interface with redaction layer
OpenRAG	Integrated	Self-hosted Milvus + BGE-small embeddings on OpenRAG VM
LiteLLM gateway	Integrated	OpenAI-compatible proxy; policy-gated cloud routing
Nero-Camp	Integrated	Workflow/approval state layer; HARP feeds evidence only
n8n	Integrated	Webhook fan-out, alert routing, low-risk event glue
Prometheus + Alertmanager	Integrated	Alert signal source; HARP reads but does not write

Enterprise Mapping

How this maps to enterprise platform engineering.

Each homelab component maps to a recognised enterprise capability. The homelab is the proof of concept; the concepts are production-grade.

HARP Component	Enterprise Equivalent	Value
HARP search (BM25 + RRF)	Enterprise RAG retrieval layer	Grounded context for AI agents and on-call engineers
MCP server (read-only)	API gateway / agent contract	Safe, scoped access surface for AI clients
Evidence packs + citations	Audit-grade answer grounding	Replaces ungrounded LLM output with traceable docs
Action tiers (R/P/X/D)	Policy-as-code guardrails	Maps to enterprise change-management tiers
Nero-Camp approval loop	Workflow orchestration + ITSM	Human-in-the-loop control plane for risky actions
Eval harness + golden corpus	Model/system evaluation	Regression testing for retrieval quality
OpenRAG connector	Vector store integration	Pluggable semantic backend behind HARP API
Redaction layer	Data-loss prevention (DLP)	Strips secrets before export to external RAG backends