Promptise Foundry

The first full-stack Agentic Engineering framework.

Build agents and the tools they use. Design how they reason. Run them as autonomous, governed systems.
Ship them to real customers — multi-tenant, secure, and observable. One framework, not a dozen libraries.

Website · Documentation · Quick Start · Blog · Discussions

## What Promptise is Promptise is **one framework for the whole job of building with AI agents** — the agents, the tools they use, the reasoning behind them, the runtime that keeps them running, and the security and governance to put them in front of customers. Not a single feature, but the full stack you'd otherwise assemble from a dozen separate libraries. Most agent stacks are assembled by hand: a model SDK, a tool layer, a vector database, auth, guardrails, a job runner, logging — glued together and kept alive by you. **Promptise pulls all of it into one framework.** `build_agent()` and a Python decorator give you the agent and its tools; memory, security, multi-tenancy, human approvals, a runtime, and observability are already inside, each switched on with a parameter. The impact: you build what your agent *does*, not the ten libraries underneath it. A prototype becomes something you can put in front of paying customers without rebuilding the production layer each time — and the same install that runs one agent on your laptop runs a fleet serving real users.
##

Get started in 30 seconds

```bash pip install promptise ``` ```python import asyncio from promptise import build_agent, PromptiseSecurityScanner, SemanticCache from promptise.config import HTTPServerSpec from promptise.memory import ChromaProvider async def main(): agent = await build_agent( model="openai:gpt-5-mini", servers={ "tools": HTTPServerSpec(url="http://localhost:8000/mcp"), }, instructions="You are a helpful assistant.", memory=ChromaProvider(persist_directory="./memory"), # remembers across calls guardrails=PromptiseSecurityScanner.default(), # blocks injection, redacts PII cache=SemanticCache(), # serves similar queries instantly observe=True, # traces every step ) result = await agent.ainvoke({ "messages": [{"role": "user", "content": "What's the status of our pipeline?"}] }) print(result["messages"][-1].content) await agent.shutdown() asyncio.run(main()) ```

_{One call. The agent finds its tools on the MCP server on its own. Memory, guardrails, cache, and tracing are each a single line —

and the ones you don't pass cost you nothing. Works with OpenAI, Anthropic, Gemini, or a local model via Ollama.}

The five parts of the framework

_{Each replaces a stack of libraries you'd otherwise wire together yourself.}

01 _🤖	### Agent One function turns any model into a working agent. `build_agent()` connects to your tool servers, discovers the tools on its own, and gives the agent what it needs to be useful in practice: memory that's searched before every reply, a security scanner that blocks prompt injection and redacts PII, response caching, sandboxed code execution, and full tracing. Each is one parameter. Any model — OpenAI, Anthropic, Gemini, a local model, or anything built on LangChain. [Agent docs →](https://docs.promptise.com/core/agents/building-agents/)

02 _🧠	### Reasoning Engine Decide how the agent thinks — or use a preset. Most tasks run fine on the default tool loop. When you need more control, lay out the agent's reasoning as a graph you can read and change: think, use tools, check its own answer, then respond. Seven presets cover common shapes — research, debate, plan-act-reflect, one-shot self-verify, write-one-program — and you build your own when none fit. No black box. [Reasoning docs →](https://docs.promptise.com/core/engine/)

03 _🔧	### MCP Server SDK Build a tool once; every agent can use it. Write a Python function, add `@server.tool()`, and it becomes an MCP tool with a schema taken straight from your type hints. The same tool works with Promptise agents and with Claude Desktop, Cursor, and any other MCP client. It comes with authentication, per-tool permissions, rate limits, circuit breakers, tamper-evident audit logs, a background job queue, and a test client that runs the whole request path without a network. [MCP docs →](https://docs.promptise.com/mcp/)

04 _⚡	### Agent Runtime Keep agents running, on budget, and recoverable. Turn an agent into a long-running process that wakes on a schedule, a webhook, or a file change. It writes down every step, so a crash resumes from where it stopped instead of starting over. Set limits on tool calls and spend, watch for stuck or looping behavior, and require a human when it hits something risky. Run one, or a fleet across machines. [Runtime docs →](https://docs.promptise.com/runtime/)

05 _✨	### Prompt Engineering Prompts you can version and test, not strings you paste. Assemble a system prompt from typed blocks with a token budget, let it change across the phases of a conversation, and check it with the same kind of tools you use for code. Version prompts, roll back a bad one, and trace exactly how each was built — so a prompt change is a reviewable diff, not a mystery. [Prompts docs →](https://docs.promptise.com/prompting/)

Built for putting agents in front of customers

_{The parts most teams end up hand-building. Here they're in the framework, off by default, on with a parameter.}

- **Multi-tenant, by construction.** Tag a request with a tenant, and every place data lives — [memory](https://docs.promptise.com/core/memory/), [cache](https://docs.promptise.com/core/cache/), [conversations](https://docs.promptise.com/core/conversations/), rate limits, [audit](https://docs.promptise.com/mcp/server/observability/) — stays separated per customer. Two customers who both have a user named `alice` can never see each other's data. It's a structural rule, not a filter you have to remember on every query. → [Multi-Tenant Platform guide](https://docs.promptise.com/guides/secure-multi-tenant-platform/) - **Human approval, enforced on the server.** Mark a tool as needing sign-off and the approval is required no matter which app calls it — including one you didn't write. Denies on timeout, rejects self-approval, records who approved what. → [Approval Gates](https://docs.promptise.com/mcp/server/approval-gates/) - **A real identity for each agent.** Agents authenticate as themselves to the APIs they call, backed by Microsoft Entra ID, AWS, Google Cloud, SPIFFE, or plain OIDC — so you can retire the shared API key, and every action traces to the person it acted for, even across agents calling agents. → [Agent Identity](https://docs.promptise.com/identity/overview/) - **Audit you can hand to a reviewer.** Every action is written to a tamper-evident chain, tied to the tenant and the user. Delete one customer's data with a single call when they ask. → [Auth & Security](https://docs.promptise.com/mcp/server/auth-security/) - **Runs offline.** The security models, embeddings, and vector store can all run locally — so the whole stack works air-gapped, for on-prem or regulated customers who can't send data out. → [Guardrails](https://docs.promptise.com/core/guardrails/) · [Model Setup](https://docs.promptise.com/getting-started/model-setup/)
##
## Everything Promptise ships _{Every capability, grouped by pillar and linked to its docs. Six parts, one framework.}

#### 🤖 Agent _{One function turns any model into a production agent.} [Explore →](https://docs.promptise.com/core/)	`Setup` [Build](https://docs.promptise.com/core/agents/building-agents/) · [Server config](https://docs.promptise.com/core/agents/server-specs/) · [Network server](https://docs.promptise.com/core/agents/network-server/) · [SuperAgent files](https://docs.promptise.com/core/agents/superagent-files/) · [Custom patterns](https://docs.promptise.com/core/agents/reasoning-patterns/) · [Cross-agent](https://docs.promptise.com/core/agents/cross-agent/) `Memory & state` [Memory](https://docs.promptise.com/core/memory/) · [RAG](https://docs.promptise.com/core/rag/) · [Conversations](https://docs.promptise.com/core/conversations/) · [Semantic cache](https://docs.promptise.com/core/cache/) · [Context engine](https://docs.promptise.com/core/context-engine/) `Security` [Guardrails](https://docs.promptise.com/core/guardrails/) · [Approval](https://docs.promptise.com/core/approval/) · [Auto-approval](https://docs.promptise.com/core/approval-classifier/) · [Sandbox](https://docs.promptise.com/core/sandbox/) `Performance` [Tool optimization](https://docs.promptise.com/core/tool-optimization/) · [Fallback](https://docs.promptise.com/core/fallback/) · [Adaptive strategy](https://docs.promptise.com/core/adaptive-strategy/) `Execution` [Streaming](https://docs.promptise.com/core/streaming/) · [Events](https://docs.promptise.com/core/events/) · [Observability](https://docs.promptise.com/core/observability/) `Reference` [Config](https://docs.promptise.com/core/config/) · [Types](https://docs.promptise.com/core/types/) · [Default prompt](https://docs.promptise.com/core/default-prompt/) · [Callbacks](https://docs.promptise.com/core/callback-handler/) · [Tools](https://docs.promptise.com/core/tools/) · [Env resolver](https://docs.promptise.com/core/env-resolver/) · [Exceptions](https://docs.promptise.com/core/exceptions/) · [CLI](https://docs.promptise.com/core/cli/)
#### 🧠 Reasoning Engine _{Reasoning as a graph you can read and change.} [Explore →](https://docs.promptise.com/core/engine/)	`Graph` [Overview](https://docs.promptise.com/core/engine/) · [Nodes](https://docs.promptise.com/core/engine-nodes/) · [Edges](https://docs.promptise.com/core/engine-edges/) · [Flags](https://docs.promptise.com/core/engine-flags/) · [Internals](https://docs.promptise.com/core/engine-internals/) `Patterns & skills` [Prebuilt patterns](https://docs.promptise.com/core/engine-prebuilts/) · [Skills](https://docs.promptise.com/core/engine-skills/) · [Skill registry](https://docs.promptise.com/core/skill-registry/) · [Custom reasoning](https://docs.promptise.com/guides/custom-reasoning/) `Runtime` [Tool injection](https://docs.promptise.com/core/engine-tools/) · [Processors](https://docs.promptise.com/core/engine-processors/) · [Hooks](https://docs.promptise.com/core/engine-hooks/) · [Serialization](https://docs.promptise.com/core/engine-serialization/)
#### 🔧 MCP Server & Client _{Build a tool once; every agent can use it.} [Explore →](https://docs.promptise.com/mcp/)	`Server` [Guide](https://docs.promptise.com/guides/production-mcp-servers/) · [Fundamentals](https://docs.promptise.com/mcp/server/building-servers/) · [Routers & middleware](https://docs.promptise.com/mcp/server/routers-middleware/) · [Auth & security](https://docs.promptise.com/mcp/server/auth-security/) · [Multi-tenancy](https://docs.promptise.com/mcp/server/multi-tenancy/) · [Approval gates](https://docs.promptise.com/mcp/server/approval-gates/) · [Production](https://docs.promptise.com/mcp/server/production-features/) · [Caching](https://docs.promptise.com/mcp/server/caching-performance/) · [Observability](https://docs.promptise.com/mcp/server/observability/) · [Resilience](https://docs.promptise.com/mcp/server/resilience-patterns/) · [Queue](https://docs.promptise.com/mcp/server/queue/) · [Advanced](https://docs.promptise.com/mcp/server/advanced-patterns/) · [Deployment](https://docs.promptise.com/mcp/server/deployment/) · [Testing](https://docs.promptise.com/mcp/server/testing/) `Client` [Guide](https://docs.promptise.com/mcp/client/) · [Tool adapter](https://docs.promptise.com/mcp/client/tool-adapter/)
#### ⚡ Agent Runtime _{Run agents unattended, on budget, recoverable.} [Explore →](https://docs.promptise.com/runtime/)	`Core` [Processes](https://docs.promptise.com/runtime/processes/) · [Orchestration API](https://docs.promptise.com/runtime/api/) · [Manager](https://docs.promptise.com/runtime/runtime-manager/) · [Context & state](https://docs.promptise.com/runtime/context/) · [Lifecycle](https://docs.promptise.com/runtime/lifecycle/) · [Hooks](https://docs.promptise.com/runtime/hooks/) · [Conversation](https://docs.promptise.com/runtime/conversation/) `Governance` [Mission](https://docs.promptise.com/runtime/governance/mission/) · [Budget](https://docs.promptise.com/runtime/governance/budget/) · [Health](https://docs.promptise.com/runtime/governance/health/) · [Secrets](https://docs.promptise.com/runtime/governance/secrets/) `Triggers` [Overview](https://docs.promptise.com/runtime/triggers/) · [Cron](https://docs.promptise.com/runtime/triggers/cron/) · [Event & webhook](https://docs.promptise.com/runtime/triggers/event-webhook/) · [File watch](https://docs.promptise.com/runtime/triggers/file-watch/) `Journal & recovery` [Overview](https://docs.promptise.com/runtime/journal/) · [Backends](https://docs.promptise.com/runtime/journal/backends/) · [Replay](https://docs.promptise.com/runtime/journal/replay/) · [Rewind](https://docs.promptise.com/runtime/journal/rewind/) `Config & scale` [Options](https://docs.promptise.com/runtime/configuration/) · [Manifests](https://docs.promptise.com/runtime/manifests/) · [Meta-tools](https://docs.promptise.com/runtime/meta-tools/) · [Coordinator](https://docs.promptise.com/runtime/distributed/coordinator/) · [Discovery](https://docs.promptise.com/runtime/distributed/discovery-transport/) · [Dashboard](https://docs.promptise.com/runtime/dashboard/) · [CLI](https://docs.promptise.com/runtime/cli/)
#### 🔐 Agent Identity _{An authenticated identity for every agent.} [Explore →](https://docs.promptise.com/identity/overview/)	`Core` [Overview](https://docs.promptise.com/identity/overview/) · [Quickstart](https://docs.promptise.com/identity/quickstart/) · [Guide](https://docs.promptise.com/identity/guide/) · [Architecture](https://docs.promptise.com/identity/architecture/) · [Security](https://docs.promptise.com/identity/security/) · [Migration](https://docs.promptise.com/identity/migration/) `Providers` [Microsoft Entra ID](https://docs.promptise.com/identity/providers/entra/) · [AWS IAM](https://docs.promptise.com/identity/providers/aws/) · [Google Cloud](https://docs.promptise.com/identity/providers/gcp/) · [SPIFFE / SPIRE](https://docs.promptise.com/identity/providers/spiffe/) · [Generic OIDC](https://docs.promptise.com/identity/providers/oidc/)
#### ✨ Prompt Engineering _{Prompts built like software — versioned and tested.} [Explore →](https://docs.promptise.com/prompting/)	`Build` [PromptBlocks](https://docs.promptise.com/prompting/blocks/) · [ConversationFlow](https://docs.promptise.com/prompting/flows/) · [Builder](https://docs.promptise.com/prompting/builder/) · [Loader & templates](https://docs.promptise.com/prompting/loader-templates/) · [Shell injection](https://docs.promptise.com/prompting/shell-interpolation/) `Strategies` [Strategies](https://docs.promptise.com/prompting/strategies/) · [Chaining](https://docs.promptise.com/prompting/chaining/) · [Context & variables](https://docs.promptise.com/prompting/context/) `Quality` [Guards](https://docs.promptise.com/prompting/guards/) · [Inspector](https://docs.promptise.com/prompting/inspector/) · [Testing](https://docs.promptise.com/prompting/testing/) · [Suite & registry](https://docs.promptise.com/prompting/suite-registry/)

_{Also in the docs}

_{📚 Guides & Labs}	[Building agents](https://docs.promptise.com/guides/building-agents/) · [Context lifecycle](https://docs.promptise.com/guides/context-lifecycle/) · [Code-action](https://docs.promptise.com/guides/code-action/) · [Production MCP servers](https://docs.promptise.com/guides/production-mcp-servers/) · [Agentic runtime](https://docs.promptise.com/guides/agentic-runtime/) · [Prompt engineering](https://docs.promptise.com/guides/prompt-engineering/) · [Multi-user systems](https://docs.promptise.com/guides/multi-user-systems/) · [Agent-to-MCP identity](https://docs.promptise.com/guides/multi-user-identity/) · [Secure multi-tenant platform](https://docs.promptise.com/guides/secure-multi-tenant-platform/) · [Multi-agent coordination](https://docs.promptise.com/guides/multi-agent-teams/) • Labs: [Customer support](https://docs.promptise.com/guides/lab-customer-support/) · [Data analysis](https://docs.promptise.com/guides/lab-data-analysis/) · [Code review](https://docs.promptise.com/guides/lab-code-review/) · [Pipeline observer](https://docs.promptise.com/guides/lab-pipeline-observer/)
_{📖 API reference}	[Agent](https://docs.promptise.com/api/agent/) · [Config](https://docs.promptise.com/api/config/) · [Memory](https://docs.promptise.com/api/memory/) · [RAG](https://docs.promptise.com/api/rag/) · [Sandbox](https://docs.promptise.com/api/sandbox/) · [Observability](https://docs.promptise.com/api/observability/) · [Identity](https://docs.promptise.com/api/identity/) · [MCP server](https://docs.promptise.com/api/mcp-server/) · [MCP client](https://docs.promptise.com/api/mcp-client/) · [Prompts](https://docs.promptise.com/api/prompts/) · [Runtime](https://docs.promptise.com/api/runtime/) · [Cross-agent](https://docs.promptise.com/api/cross-agent/) · [SuperAgent](https://docs.promptise.com/api/superagent/) · [Utilities](https://docs.promptise.com/api/utilities/)
_{🚀 Start here}	[Installation](https://docs.promptise.com/) · [Extras](https://docs.promptise.com/getting-started/installation-extras/) · [Quick start](https://docs.promptise.com/getting-started/quickstart/) · [Cookbook](https://docs.promptise.com/getting-started/cookbook/) · [Why Promptise](https://docs.promptise.com/getting-started/why-promptise/) · [What is MCP?](https://docs.promptise.com/getting-started/what-is-mcp/) · [Model setup](https://docs.promptise.com/getting-started/model-setup/) · [Best LLMs](https://docs.promptise.com/getting-started/best-llms-for-agents/) · [Key concepts](https://docs.promptise.com/getting-started/concepts/) · [Glossary](https://docs.promptise.com/getting-started/glossary/) • More: [Blog](https://docs.promptise.com/blog/) · [Showcase](https://docs.promptise.com/resources/showcase/) · [Examples](https://docs.promptise.com/resources/examples/) · [Migration](https://docs.promptise.com/resources/migration/) · [Changelog](https://docs.promptise.com/resources/changelog/) · [FAQ](https://docs.promptise.com/faq/) · [Contributing](https://docs.promptise.com/resources/contributing/)

Ecosystem

_{Promptise plugs into what your team already runs — and runs fully offline when it has to.}

#### Models

_{+ any LangChain BaseChatModel · FallbackChain for automatic failover · Model setup →}

#### Memory & Vectors

_{Local embeddings · air-gapped model paths · per-tenant isolation · Memory →}

#### Conversation Storage

_{Session ownership enforced · per-tenant isolation for cache and guardrails · Conversations →}

#### Identity & Auth

_{A verifiable identity per agent — no shared keys · Agent Identity →}

#### Observability

_{8 transporters: OTel · Prometheus · Slack · PagerDuty · Webhook · HTML · JSON · Console · Observability →}

#### Sandbox & Deployment

_{Docker + seccomp + gVisor + capability dropping · Kubernetes health probes · Sandbox →}

#### Protocols

_{stdio · streamable HTTP · SSE · HMAC-chained audit logs}

_{Want to ship with us? See CONTRIBUTING.md · join Discussions · file an issue.}

---

[**Contributing**](CONTRIBUTING.md) · [**Security**](SECURITY.md) · [**License: Apache 2.0**](LICENSE)

_{Built by Promptise}

_{^{Formerly DeepMCPAgent — a public preview of one sliver of this framework (MCP-native agent tooling). Promptise Foundry is the full system it hinted at: reasoning engine, agent runtime, prompt engineering, sandboxed execution, governance, and observability.}}

Promptise Foundry

Get started in 30 seconds

The five parts of the framework

01

02

03

04

05

Built for putting agents in front of customers

Ecosystem