Memory and recall
Agent Brain
Self-hosted shared memory, semantic cache, and governance for Claude, Codex, Cursor, Gemini, Aider, and custom agents.
Agent Software Suite
A cohesive suite for developers building with AI agents: shared memory, voice capture, multi-agent orchestration, and a terminal built around assisted development.
Live product surfaces across the suite: Agent Brain, Agent Wispr, Agent Harness, Agent Terminal, and Agent Studio.
wispr.capture("ship the harness rollout notes")
brain.remember(decision)
harness.assign(workerTask)
terminal.run(verification)
// one workflow across the suiteProducts
Each product has a distinct role. The suite hub keeps status, docs, and updates in one place while product-specific docs stay with each product.
Memory and recall
Self-hosted shared memory, semantic cache, and governance for Claude, Codex, Cursor, Gemini, Aider, and custom agents.
Voice capture
Voice-first capture for agentic development notes, prompts, decisions, and task handoffs.
Agent orchestration
AI agent execution harness with local models, verification, approvals, routing, and recovery.
Developer terminal
Voice-to-agent developer terminal. Speak a goal, enrich with Brain memory, run the pre-send guard, and delegate to Harness workers — from one native desktop app. Log in once and stay signed in across the full suite.
Launch video studio
Done-for-you, studio-grade SaaS launch videos — real app screen capture, pro voiceover, a 3D logo intro, and music. The presenter can be your own AI-cloned voice or face.
How it fits
The suite is organized around the actual lifecycle of agentic work: capture context, remember decisions, coordinate bounded tasks, and operate from a local development workspace.
Agent Wispr
Turn spoken notes, debugging context, and implementation intent into structured material agents can use.
Agent Brain
Persist durable project knowledge and retrieve relevant context across tools and sessions.
Agent Harness
Assign bounded tasks, verify expected changes, and keep multi-agent implementation work disciplined.
Agent Terminal
Bring local commands, project context, and assisted development flows into a terminal built for the suite.
Self-hosted
Agent Brain, Agent Harness, and Agent Terminal are self-hosted. No data leaves your environment unless you explicitly push it somewhere.
Local models via Ollama or LM Studio, local vector storage, local history. You choose which providers you call — the suite works without any external API.
Agent Brain stores context in a database you control. Nothing is shared with Anthropic, OpenAI, or any third party. Recall and forget on your own terms.
Authenticate once and your session carries across Brain, Terminal, and Harness automatically. Credentials are stored in the OS keychain — never in plain files.
Pricing
Individual Pro plans from $19/mo. The All-Access Pass covers all four at $59/mo.
Updates
The new Agent Software Suite hub gives builders one place to track live products, in-development tools, documentation links, release notes, and future bundle availability.
By Agent Software
Agent Brain's product-site work set a practical bar for the suite hub: clear status, careful checkout rollout, documentation links, metadata, feeds, and repeatable verification.
By Agent Software
Affiliate Program
Refer paying builders and earn 30% on every invoice they pay for the next 12 months.
30%
Commission
12 invoices
Window
60 days
Cookie
Join Discord, browse product docs, or follow the feed for release notes and readiness updates.
Yes. The All-Access Pass ($59/mo or $590/yr) covers Agent Brain, Agent Wispr, Agent Terminal, and Agent Harness under one subscription. Individual product plans start at $19/mo.
Agent Brain, Agent Wispr, Agent Harness, and Agent Terminal each have dedicated product sites with docs, download, and roadmap pages. Agent Studio — done-for-you SaaS launch videos — is also live.
This site keeps a suite documentation directory and links to each product's own docs when those docs are available.
Use the blog, RSS feed, JSON feed, Discord, or email support to follow release-readiness and product updates.