---
title: Public mesh hardening
description: How the public mesh checks that a daemon run by a stranger serves the registered model at the promised quality, what was measured, and what is still weak.
section: reference
order: 180
audience: [operator, builder]
sources: [docs/public-mesh-hardening.md, docs/public-mesh.md, services/go-orchestrator/internal/audit, services/go-orchestrator/internal/provenance, services/vllm-venom-sampler]
features: [reference.doc-public-mesh-hardening]
---

On the public mesh, anyone can host a daemon and earn per token. This page summarises how venom checks such a daemon and where the checks are still weak. The normative contract is `docs/public-mesh.md` and the decision record is ADR 0023; where they differ from this summary, they win.

## The threat {#threat}

A public daemon's operator can earn the same money more cheaply by serving a smaller model, a coarser quantisation or a cheaper engine setting (an FP8 KV cache); by editing, truncating or caching prompts and answers; by inflating token counts; by faking latency to attract traffic; or by behaving only when it thinks it is tested. The daemon's code is open, so no check may depend on a secret the daemon can see.

## The pipeline {#pipeline}

1. **Verifiable sampling (`venom-sampler/1`).** Every public request carries a secret 16-byte nonce. The engine picks each token as the Gumbel-max over the top-k/top-p candidates, with noise from `HMAC-SHA256(nonce, position ‖ token)`. Given the model's log-probabilities, anyone can replay the choice. A vLLM logits-processor plugin implements it.
2. **Receipts.** The daemon signs a receipt binding the prompt token hash, the output token ids, the streamed text and the counts.
3. **Audits.** An auditor replays sampled jobs on reference hardware and compares the daemon's tokens with the ones the registered model would pick.
4. **Sequential evidence.** One audit is tuned to a one-in-a-million false-positive rate, which makes it blind to an artifact that is only slightly wrong. The auditor therefore accumulates evidence across audits (a CUSUM-style statistic), so a small, steady excess, such as a one-step-coarser quantisation, is caught over many jobs.
5. **Witnessed latency.** The cloud stamps time to first token on the first streamed line that carries a real output token, so an empty opening chunk sent at once no longer looks fast. A "first-token feint" (streaming a guessed opener before prefill ends, then sampling honestly) is caught by the sequential evidence at its second audited occurrence.
6. **Preflight.** A registry entry can carry up to 8 golden cases. At start-up and every 6 hours the daemon runs them through its public-job path, locally, and advertises a model for public traffic only when they pass. The cloud never sees a preflight.
7. **llama.cpp without a plugin.** A shim (`venom-llamacpp`) puts llama.cpp behind the sampler contract so it can be audited too.

## Known shortfalls {#shortfalls}

- Verdict results remain visible to the daemon after each epoch, so a daemon learns that it failed an audit, though not how close it is to the alarm.
- A non-streamed job's time to first token is its whole duration; a stream padded with over 64 KiB of tokenless data falls back to the done time and counts as an SLO miss, not as abuse.
- The preflight's thresholds come from a toy model and have not been run against real vLLM or llama.cpp.
- The llama.cpp path replays every prompt position, which is slow on long prompts, and approximates the rank of rare prompt tokens.
- Region verification, an attested trust class, payouts and KYC are out of scope.

## Reproducing {#reproducing}

In a venom checkout, the adversarial toy suite is `go test ./internal/audit -run Adversarial -v` in `services/go-orchestrator`, and the shared vectors run in `go test ./internal/provenance -run Cusum` and `make test-cloud`. `make public-mesh-e2e` runs the whole public mesh on loopback.

## Related

- [Glossary](/docs/reference/glossary.md)
- [Design decisions (ADRs)](/docs/reference/adr.md)
