Skip to content
View bulletproofsoftware-ai's full-sized avatar

Block or report bulletproofsoftware-ai

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
.github/profile/README.md

bulletproofsoftware-ai

Open-source infrastructure for building reliable, governed, self-hostable AI-agent systems.

A suite of composable, self-hosted building blocks for AI agents — persistent memory, multi-agent orchestration, governance and runtime security, an agent data plane, and the developer tooling that ties them together — plus a working assistant assembled from them. Everything runs on your own infrastructure.

🌐 bulletproofsoftware.ai


🧠 Memory

Repo What it does
bulletproof-memory Persistent, tiered memory for AI agents — an MCP server with semantic recall over Qdrant + Postgres. Includes the web dashboard and the turnkey stack.
bulletproof-process-knowledge-mcp MCP server exposing a process-knowledge Qdrant collection: semantic query, exact lookup, record validation.

🎼 Orchestration

Repo What it does
bulletproof-conductor-kernel Domain-agnostic multi-agent orchestration kernel: tier classification, agent dispatch, gate enforcement, state persistence.
bulletproof-conductor-dev The dev-domain workflow orchestrator built on the kernel — tiered pipelines with programmatically enforced verification gates.
bulletproof-event-router Central event router: validate, route, and dispatch events to n8n webhooks, state, and actions, with a dead-letter queue.
bulletproof-metrics-engine Outcome metrics and predictive scaling from an event stream.

🛡️ Governance, Security & Compliance

Repo What it does
bulletproof-governance-plugin Agent governance framework: identity manifests, trust broker, audit bus, policy engine, memory governor.
bulletproof-runtime-security Runtime security + SOC2 evidence for agents: identity, behavioral monitoring, threat detection, memory integrity.
bulletproof-codehardener Code assurance platform: multi-tool security scanning, quality scoring, and hardening (MCP + dashboard).
bulletproof-validation-gate A Claude Code Stop hook that adversarially verifies completion claims and blocks unsupported "done" claims.
bulletproof-compliance-portal Compliance & governance portal: DSR management, audit explorer, evidence packages, gate decisions, regulatory reports.
bulletproof-compliance-service Compliance backend: source-of-truth compliance data, JWKS signing, tamper-evident audit chain.
bulletproof-compliance-jobs Scheduled compliance jobs: data retention, DSR cascades, Merkle-chained audit signing.

🔌 Agent Infrastructure

Repo What it does
bulletproof-a2a-gateway Agent-to-agent gateway: authenticated, rate-limited, audited capability invocation with an MCP bridge.
bulletproof-agentic-data-plane A governed data plane for agents: GraphQL + REST with classification, access control, and audit.
bulletproof-conductor-data-pipeline Data pipeline with masking/anonymization, JSON-schema contracts, lineage, and human approval gates.
bulletproof-agent-economics Cost metering, budget enforcement, model routing, and chargeback for multi-agent systems.

🤖 Assistants

Where the building blocks above are assembled into something you can actually talk to.

Repo What it does
bulletproof-monet A self-hostable personal AI assistant reachable over Telegram: tiered memory, a prompt-injection gate and audit trail, a webhook receiver, background jobs, and an opt-in daily briefing. Bundles the memory MCP server; voice output uses Microsoft edge-tts (no API key).

🧰 Tooling

Repo What it does
bulletproof-context-guard Claude Code plugin: context-window monitor with 4-tier escalation and a 5-compartment usage MCP server.
bulletproof-markdown-for-agents Self-hosted HTML→Markdown proxy that serves clean, agent-ready content.
bulletproof-prd-platform Full-stack PRD platform: brainstorm, create, and review product requirement documents.
bulletproof-design Self-hosted brand & design asset library: components, icons, guidelines, brand tokens.
bulletproof-docs-portal Minimal web app to browse and PDF-export project documentation.

Getting started

Most services are self-hostable via Docker and ship with an .env.example and a README. The memory system is a good place to start — one clone + ./install.sh brings up the full stack.

All repositories are licensed Apache-2.0.

Popular repositories Loading

  1. bulletproof-memory bulletproof-memory Public

    Persistent, tiered memory for AI agents — an MCP server with semantic recall over Qdrant + Postgres.

    TypeScript

  2. bulletproof-context-guard bulletproof-context-guard Public

    Claude Code plugin: context-window monitor with 4-tier escalation and a 5-compartment usage MCP server.

    Shell

  3. bulletproof-markdown-for-agents bulletproof-markdown-for-agents Public

    Self-hosted HTML-to-Markdown proxy that serves clean, agent-ready content.

    Python

  4. bulletproof-event-router bulletproof-event-router Public

    Central event router: validate, route, and dispatch events to n8n webhooks, state, and actions with a dead-letter queue.

    Python

  5. bulletproof-process-knowledge-mcp bulletproof-process-knowledge-mcp Public

    MCP server exposing a process-knowledge Qdrant collection: semantic query, exact lookup, and record validation.

    Python

  6. bulletproof-metrics-engine bulletproof-metrics-engine Public

    Outcome metrics and predictive scaling from an event stream (pairs with bulletproof-event-router).

    Python