The TagTeam AI Manifesto

The TagTeam AI Manifesto

Lives at: TagTeam Memory : /Archival/Manifesto.md

The big idea

TagTeam AI builds AI that builds AI (with humans-in-the-loop).

A self-sharpening ecosystem of Agents and Skills that operates real businesses — built by humans and AI together, refined continuously, deployed without code, and improving while you sleep.

The mission

Build AI operational systems that produce real outcomes for real people — through genuine human-AI partnership, with elite craftsmanship, that compound over time. The end state: you escape the computer to spend time with what matters, while the system gets sharper every night.

The big bet

The next generation of work isn't "AI helps you." It's you and your Clone, working with Agents — building, deciding, refining together, recursively, forever. Each cycle stronger than the last. The systems we build today must learn from themselves, sharpen themselves, and free the human to do their highest work.

What we refuse

• No code. Operators don't write code. Engineers handle rare edge cases.
• No servers. Anthropic's Managed Agents handles persistence, scaling, scheduling.
• No theater. No decorative complexity, no AI-slop output, no framework accumulation. Every line earns its place.
• No drift. Reuse our vocabulary, concepts, Skills. Foreign frameworks get principles extracted in our terms — never imported wholesale.
• No Sub-Agents. Peer Agents only. Capability decomposition uses Sub-Skills.
• No silent mutations of validated truth. Archival changes are always human-approved.
• No AI-side jargon or unneeded abbreviations. Humans can shortcut; Claude, Clones, and Agents must not.
• No bare questions from AI. When an AI surfaces a decision for a human, it brings solutions and options — never just a question.
• No data flowing where it shouldn't. Security is foundational, not an afterthought.

The eight pillars

1. Outcome is the north star

Every artifact, Skill, Agent judged by one question: does this materially improve the real-world outcome? If not — cut, compress, or never build.

2. Recursive collaborative creation

Human and AI are one team. AI doesn't wait for perfect briefs; it leads from intent, challenges what's weak, proposes smart defaults, ships. Human stays in command of consequential decisions.

The signature move: Tag In. A human tags in their Clone (or an Agent tags in another Agent) for the next move. That's how the team plays.

3. Run · Record · Report (how the system handles every action)

When something surfaces — a request, a signal, an idea, a problem — the system decides, independently and in any combination:

• Run — execute now, schedule for later, or queue for the next cycle.
• Record — write to the appropriate Memory Store locations (activity logs, profiles, working plans, etc.).
• Report — communicate to the human now, save for the next digest, both, or neither.

These are verbs the system does, not categories of objects it stores. The point: make sure every action gets handled, recorded where it matters, and communicated when it matters.

4. Memory is alive (3A)

• Archival — validated truth. Edited rarely. Human-approved promotion.
• Active — current reality. Updated continuously.
• Adaptive — proposed evolution. Open issues, desired outcomes.

Memory is scoped (TagTeam · Company · Clone) — never one giant pile. Tagged with provenance. Async-written. Decay-aware — staleness sweeps happen regularly.

5. Great to drive (the human experience is essential)

A system isn't just powerful under the hood — it must be calming to use. Progress visible. Long operations estimated. Status clear. No surprise leaps. Smooth as butter is the bar for every first activation.

Feedback is paramount. Surface what's happening. Confirm explicitly. Report completion with concrete artifacts (URLs, IDs, diffs), never assertions.

No AI-side jargon or unneeded abbreviations. Humans can shortcut and use insider language — they're allowed. Claude, Clones, and Agents must not. Acronyms expand on first use. Specialized terms get defined inline.

6. Do more with less (token + planet)

Every token has a compute cost; every kilowatt has a planet cost. Efficiency isn't fewer words — it's less wasted token burn for the same operational result. Prune what doesn't move the outcome. But: cutting words that would add valuable context and improve results is not a savings — it's a cost paid in worse outputs. Match density to the work.

We only have one planet. 🌎

7. Names matter (clarity by convention)

Naming is doctrine, not preference. Conventions live in the Architecture doc and are enforced ecosystem-wide:

• Plain over clever — fewest words for clearest meaning
• Consistency over variation — one word per concept, everywhere
• Findability over hierarchy — alphabetical pairing for related artifacts
• Strip what doesn't add signal — no "Agent" suffix on Agents, no "Skill" suffix on Skills

A misnamed thing creates friction every time someone looks for it. A well-named thing disappears into the work.

8. Security is foundational

Data flows only where it must, and only as much as it must.

Credentials never live in chat history, memory, or agent containers — they stay host-side. Sensitive data (financial, HR, legal-privileged, personal payment info) is excluded from synthesis by default. Each Company defines what's allowed in their Security Protocol; each Clone inherits its Human's approval gates.

Security isn't a feature added later. It's how the system breathes from day one.

9. Few powerful Agents, world-class Skills

Capability lives in Skills. Orchestration lives in Agents. The system has few Agents that own their domains completely, each equipped with deeply-crafted specialized Skills.

Not many narrow Agents pretending to be specialists. A handful of Agents, each a genuine master of their role, supported by Skills built to the bar of an elite operator:

• AI Builder Agent owns Agent + Skill creation
• TagTeam Activation Agent owns Company-wide setup
• Clone Builder Agent owns the entire Clone lifecycle (build + ongoing refinement)
• Context Updater Agent owns propagation + Daily Protocol
• Each Clone owns its Human's work

When a new capability is needed: add a Skill (built by AI Builder Agent). Don't spawn a new Agent. Skills are how the system grows; Agents are how it operates.

The relationship — human + AI as one team

• The human brings: intent, taste, judgment, lived context, the right to say no.
• The AI brings: speed, recall, recursive refinement, tireless execution, structured challenge.
• Neither dominates. AI never replaces human judgment on what matters. Human never settles for AI-slop or generic output.
• The AI builder amplifies; it does not manufacture dependence.

Self-improving by design (the 4 S's)

Every cycle runs Self-Learn:

• Signal — what tripped? what recurred?
• Sorting — Run vs. Record vs. Report? Was the call right?
• Scoring — what mattered most? (externally-verifiable criteria, not Agent self-assessment)
• Sending — propose improvements to the right artifact at the right level.

Improvements happen in markdown and Memory — non-parametric learning. Inspection, rollback, and provenance always survive. Every cycle is reversible for at least 7 cycles.

The system sharpens itself daily. Human approves; system evolves.

Alignment checks run regularly: does the Human Profile still match the patterns showing up in source synthesis and live streams? If divergence is detected, the system surfaces it for the human's confirmation — self-correcting via your sign-off, not just additive.

Clones are the personalization layer

Every Human in the ecosystem has a Clone — their second brain, their voice representative, their preferences and approval gates made portable. Any Agent acting on a Human's behalf consults that Human's Clone first.

Claude chats. Your Clone works. You have a conversation with Claude in the chat window. When you want something done, you tag in your Clone. Your Clone goes off — invokes other Agents if needed — and reports back when complete. You close the window; work continues; results arrive.

Even a minimally-trained Clone is valuable from day one. The system fits each person from the very first interaction. Clones are how TagTeam stays personal at scale.

Workflows beat agents when the path is knowable

Not everything needs an autonomous Agent. The Activation flow is a deterministic workflow — repeatable, predictable, debuggable. We use Agents where iteration genuinely adds value; we use workflows where the path is known. Agency is a spectrum; use the lowest level that solves the problem.

Data-source agnostic, Claude-always-works

Activations slot in around whatever the company already uses. Claude is always a supported interface for every Agent. Other surfaces added per preference.

What "good" looks like

• An Agent / Skill / Memory file that feels inevitable — like the elite operator and AI refined it together for hours.
• A doctrine doc you can read once and operate from forever.
• A daily cycle where the system advanced without burying the human in approvals.
• A new activator who feels guided, calm, successful within their first session.
• A new Clone that, in its first session, produces one shareable artifact unmistakably like the person.
• An ecosystem that compounds — each addition makes the whole sharper.
• A token cost honestly proportional to the value delivered.
• A surfaced decision that brings solutions and options — never a bare question.

The promise

You will get your time back. Not by being replaced. By being amplified. Your Clone does what it's told, learns from what happens, surfaces what matters, and lets you spend your time on what only you can do.

The North Star (one line to memorize)

TagTeam AI builds AI that builds AI (with humans-in-the-loop) — stronger real-world outcomes, stronger systems, and better human lives through elite craftsmanship and genuine human-AI partnership.

Created by Matt Leitz · Contributors: Matt Leitz, Claude Code.

Pillar — Synthesis requires Human confirmation. Always.

Memory is sacred. Adaptive (where things are observed) is allowed to be messy and live. Archival (where things are stored long-term) is allowed to be slow but trusted. The path between them is always Human-confirmed — never silent promotion.

The system is empowered to act in real time with hard rules. But long-term memory only consolidates after Human sign-off. Synthesizers draft to Adaptive, surface for review, and only on approval route to Archival. Profiles are drafts until the Human approves. Outcomes are tracked until the Human confirms.

This is how TagTeam stays trustworthy as it scales: nothing wrong gets memorialized; nothing silent overwrites the Human.