The foundation layer
for agentic intelligence.
Most AI agents never make it past the demo: they are unpredictable, unsafe and impossible to audit. Promptise is the open-source framework that gets them into production: agents that reason, connect safely to your systems, and run under your rules.
- Open source · Apache 2.0
- Works with any model
- Runs on your infrastructure
Everything between your model and production, in one framework.
An agent that does the work, an interface that connects it safely to your systems, and a harness that keeps it running under your rules. Most teams spend months stitching these together from separate libraries. Promptise ships all three in one install.
Agent
It thinksAgents that work through a task instead of guessing: they plan, use tools, check their own output and answer, with prompts managed like code.
Interface
It actsSafe access to your systems. Build MCP servers with authentication and audit built in, connect to anyone else's, or turn the API you already have into one an agent can use.
Harness
It operatesWhat keeps agents running and accountable: schedules and crash recovery so the work gets done, budgets and health checks so costs stay bounded, identity and audit so you can prove what happened.
See one customer request handled from start to finish.
A customer writes in at three in the morning. Follow the ticket step by step: what wakes the agent, what it knows, how it reasons, what it calls, who signs off and what it records.
Open the full walkthroughAgent Runtime
A ticket wakes the agent
A helpdesk webhook fires. The runtime wakes the support agent's process and hands it the ticket.
Agent · It thinks
Interface · It acts
Harness · It operates
An illustrative run. Names and timings are examples; what each step describes is what the framework does. See every module in the full walkthrough →
Twelve modules. Start with the one you need.
Each module solves one production problem. Add the others when you need them: they share identity, audit and configuration, so nothing has to be glued together. Pick one to see what it does.
The Promptise Agent
One function returns a production agent. Any model, tools discovered from your MCP servers, memory searched before every turn, and guardrails, caching and tracing switched on by a flag.
3
memory providers4
conversation stores4
tool-selection levels- Any model: a provider string or any LangChain chat model
- MCP tool discovery, with semantic tool selection
- Memory in Chroma, Mem0 or in-memory
- Conversations persisted to SQLite, Postgres or Redis
- Per-request caller identity through cache, guardrails and tracing
from promptise import build_agentfrom promptise.config import HTTPServerSpecfrom promptise.memory import ChromaProvider agent = await build_agent( model="openai:gpt-5-mini", servers={"crm": HTTPServerSpec(url="http://localhost:8000/mcp")}, memory=ChromaProvider(), guardrails=True, observe=True,)What it takes to move an agent from demo to production, built in.
Eight capabilities, each part of the framework rather than a plugin you add later, so they share one configuration, one identity and one audit trail.
Context
Answers grounded in what the agent knows, without overflowing the context window or the budget.
Reasoning
Reasoning designed for the job, so results are predictable instead of left to one generic loop.
Tools & integration
Connect to any system through the open MCP standard, and publish your own APIs the same way.
Safety
Prompt injection, personal data and leaked credentials are caught before the model sees them, and again on the way out.
Governance & controls
Spending limits and approvals enforced in code, not in the prompt, with a person signing off where it counts.
Autonomy
Agents run on their own schedule and pick up where they left off when something breaks.
Identity & audit
Know which agent acted, and prove it to someone who was not there.
Observability
Every model call, tool call and retry traced, and sent to the tools your team already uses.
From pip install to a running agent, in three steps.
Python 3.10 or newer. No scaffolding and no project generator: it works inside the code you already have.
$ pip install promptise # or, with vector memory, ML guardrails and exporters$ pip install "promptise[all]"Free guides on building with LLMs.
Practical articles on prompt engineering, LLM security and building with models: one technique per guide, with worked examples. Read any one on its own, or follow a learning path that puts them in order.
- 70
- Guides
- 4
- Learning paths
- 6
- Topics
What is Prompt Engineering?
Prompt engineering is the skill of shaping inputs so LLMs like ChatGPT deliver clear, accurate, and useful results. This guide walks you through the foundations, core techniques, and practical strategies to help you design prompts that truly work.
Read the guide→- 02Structuring Prompts for Safety: Layers That HoldPrompt Security · Advanced · 22 min→
- 03The LLM Attack Surface: Complete Technical Guide to Vectors, Risks, and MitigationsLLM Security · Advanced · 35 min→
- 04LLM Supply Chain Security: Dependencies, Models, and TrustLLM Security · Advanced · 15 min→
- 05The EU AI Act for LLM Builders: What It Actually MeansCompliance · Beginner · 8 min→
Questions teams ask before they adopt it.
Short answers. The docs have the long ones.
The agent framework, a production SDK for building MCP servers, a prompt-engineering layer and an agent runtime. MCPcast, which turns an existing OpenAPI document into a reviewed MCP server, ships with them.
Get your first agent into production.
Open source and free under Apache 2.0, with no account to create. Install it with pip install promptise and follow the quickstart.