Proof Lab
StartEcosystem
Explore (579)Live Systems (52)Pricing
Log InGet API Key✓ Verify It Yourself
Agent-008 · Decision Integrity Infrastructure

Was this agent allowed to do that?

Agent-008 is decision-integrity infrastructure for autonomous AI agents. It gates every agent action on certified authority and emits a Provable Authority Package — portable evidence of why an action was allowed, or a negative authority proof of the governed action that was prevented. It answers the question above with cryptographic authority evidence rather than with a policy document.

Definition

Agent-008 is a runtime gate that authorizes AI-agent actions and emits a Provable Authority Package (PAP) — portable evidence of why an action was allowed, or a negative authority proof (NAP) of the governed action that was prevented. The decision it makes is authorization, not correctness. Agent-008 is not an AI safety model — it is an authority system for AI actions. It does not own the substrates it uses, does not make model outputs correct, and does not replace your identity system.

Agent-008 does not decide whether an AI answer is correct. It decides whether the requested action is authorized.

Read the whitepaper →Read the Adversarial Testing Report → ✓ Verify it in Proof Lab

Standalone Rust binary · deterministic gate · offline-verifiable evidence

The Authority Timeline

A decision needs a track to travel on.

An autonomous system needs more than good decisions. It needs a trusted history of authority — because the question an auditor asks is rarely “what did it decide?” and almost always “what was it allowed to decide, at that moment, and who moved that boundary?”

Agent-008 preserves that chain on H33-Stamption, the post-quantum governance commitment layer, so every authorization, transition and evidence event travels on one verifiable track. The railway is the useful picture:

The track
proves where the action was allowed to go — the authority boundary as it stood at that moment, not as it was reconstructed afterwards.
The signals
prove who authorized movement, and under what delegated scope.
The receipts
prove what actually happened — the evidence the decision produced, and whether it reproduces.

The chain Agent-008 preserves runs human intent → authority → execution → evidence. Code, build and artifact identity along that chain come from H33-GIT; Agent-008 governs the runtime and does not claim code identity as its own.

Stated plainly: H33-Stamption and H33-GIT are admitted identities with no published deployment of their own yet. Agent-008’s own evidence — the whitepapers, the adversarial testing report, the offline bundle below — covers Agent-008. Composition does not extend it to what Agent-008 is built with.

The Problem

Memory is not an authority system.

Autonomous agents act from memory, prompts, and inherited context. None of that is an authority system — it is a guess about what the agent is allowed to do. The risk is not merely an incorrect answer; it is an unauthorized action: an agent moving funds, changing a policy, or calling a tool that no human ever authorized — with no way to prove afterward whether it was allowed.

How It Works

A deterministic gate. No model in the loop.

Every action passes a deterministic runtime gate. The gate is a boolean evaluation over signed inputs — so it is replayable and its verdict is stable.

1
Human intent root
Authority traces back to a certified root of human intent (H33-Root). An action that cannot cite a certified Root is denied.
2
Delegation & authority state
Delegation and scope are governed by Q-Sign; the gate checks the action against the current authority state.
3
Decision gate
Four deterministic authorization checks run; every denial identifies the failed authorization condition (e.g. no_read_attestation, capsule_invalid, not_activated, stale_authority). No silent allow, no silent reject.
4
PAP / NAP evidence
The decision is sealed as a PAP (allowed) or NAP (prevented), carrying intent, root, policy, attestation, and lifecycle.
5
Replay & revocation
The decision is replayable later; captured secrets can be revoked. Verification survives even if H33 disappears.
Composition

Agent-Zero is a part, not the whole.

Agent-008 composes Agent-Zero as its privacy-preserving classification layer — Agent-Zero enables privacy-preserving classification workflows on CKKS ciphertexts, so protected data can stay encrypted through the computations it supports. Agent-Zero is a distinct product with its own identity; Agent-008 uses it as a part, it does not absorb it.

Boundaries

What it depends on — and what it does not own.

Agent-008 is built on shared substrates it uses — it never claims to own them:

Relationship
Concept
Role
DEPENDS_ON
Q-Sign
delegation & authority governance
USES
H33-Root
certified root of human intent
USES
Verification (ZK-Verify)
independent artifact verification
USES
H33-Key
secret protection (raw values never reach the agent)
USES
H33-Upstream
provenance & lineage
USES
Agent-Zero
privacy-preserving agent operations
USES
TFHE
computation on encrypted data
BUILT_ON
H33-74
the compact post-quantum evidence primitive used to bind and replay governance events — what the PAP is sealed on
COMPOSES
H33-GIT
code, build and artifact identity for the software that executed (admitted identity; no published deployment yet)
COMPOSES
H33-Stamption
the governance commitment layer the authority chain is preserved on (admitted identity; no published deployment yet)
MONITORED_BY
HATS
continuous evidence of governance (external — not a component Agent-008 owns)
The Operating Lifecycle

Agent-008 is the govern step.

APQC migrates an environment to a governed state, Agent-008 enforces authority on every action, HATS continuously monitors that it holds, Verification proves it independently, and Authority Center preserves the resulting authority over time.

APQC migrate Agent-008 govern HATS monitor Verification prove Authority Center preserve
What Proves It

Re-verify the decision yourself. Offline.

The strongest evidence is a downloadable audit bundle you re-verify offline, with no contact to H33. The proof page is not the proof — the artifact is.

The website is not the proof. The artifact is the proof.

Offline verification bundle

A signed evidence tar with signer_pub.bin and an offline verification_report.md — re-check the decision yourself, and confirm a substituted value is rejected.

Run the offline verifier →
PAP / NAP artifacts

Each decision seals a Provable Authority Package (allowed) or negative authority proof (prevented), verifiable independently by a zero-state verifier.

Preflight a decision →
Developer Path

One command. An authorized decision plus its PAP.

Point Agent-008 at a request and it returns an authorized decision plus its Provable Authority Package:

$ agent-008 run --request <path>

A local HTTP serve mode is coming next — disclosed here as not-yet-available rather than presented as shipped. Documentation · Whitepaper

The Category Difference

Not a stronger control. A different kind of one.

Agent-008 does not ask whether an AI is intelligent enough to act. It asks whether that action had cryptographic authority to occur.

The contrast below is with a category, not with a vendor. Conventional AI controls describe intent and record activity; both are useful and neither produces something an outsider can check.

Conventional AI controls
Agent-008
Policies
Cryptographic authority
Permissions
Governed decisions
Logs
Replayable evidence
Model evaluation
Action authorization
Trust the system
Verify the artifact

The right-hand column is not a claim that the left-hand one fails. It is a claim about what each produces: a policy states what should happen, a log states what was observed, and an authority proof states what was permitted — in a form somebody outside the system can check without asking us.

Who Needs It

Agents that take real actions.

Teams deploying autonomous agents that move value, change state, or call tools. Agent-008 replaces prompt- and memory-based “trust the agent” with a deterministic, provable authorization gate. Especially where an unauthorized action carries financial, operational, regulatory or safety consequences.

Without it: an agent can take an action no human authorized, and you cannot prove afterward whether it was allowed.

See Agent-008 govern your agents
Limitations & Boundaries

What it does not do.

Agent-008 does not make model outputs correct, does not replace an identity provider, and does not prove any claim without an artifact. It does determine whether an action is authorized and preserve replayable evidence of that decision.

In progress · disclosed, not asserted as live

Authority Freshness enforcement and Root-lineage verification are on the roadmap; the current verification level reports stage_d_attest_bound and advances to aggregate_v1 when Root lineage wires in — not a fake green check.

Research access

Test Agent-008

Apply for research access to run governed adversarial testing and independently review the resulting evidence and Replay.

Both options record an application only. Neither creates a licence, entitlement, activation or credential.

This form records an application for review. Submitting it does not grant access to Agent-008, and does not create entitlement, activation or execution authority. We collect only what is needed to adjudicate the request — no credentials and no payment details.

Related products

Products built on Agent-008

H33-RootAgent-ZeroQ-SignAI DefenseAI DetectionAI Governance