Omnis LocusTrace · formerly PatentChecker

Verifiable patent-sequence drift monitoring

Omnis LocusTrace produces offline-verifiable evidence packets showing how public patent-sequence coverage changes around a biotech program over time.

Informational evidence infrastructure only. Not legal advice, not infringement analysis, not FTO.

Start with a bounded watch area and get reviewable evidence packets, not screenshots or dashboard exports.

LocusTrace WatchlistsMaterial deltasDeclared scopeLocusTrace Evidence Packets
Default cadence
Weekly+

Weekly monitoring by default, with tighter cadence for high-volatility programs.

Core outputs
3 artifacts

Human digest, machine alert summary, and a counsel packet index with hashes.

Verification mode
Offline

Reviewers can verify packet integrity without source access or registry credentials.

Money shot
Omnis LocusTrace output showing patent-sequence drift after baseline with deterministic Evidence Packet structure and risk change summary.
Representative output from a deterministic demo packet. Informational evidence infrastructure only. Not legal advice, not infringement analysis, not FTO.
Receipt snippet
{
  "packet_fileset_hash": "sha256:<...>",
  "deterministic": true,
  "artifacts": {
    "packet_zip": "demo_offline_packet.zip",
    "packet_zip_sha256": "demo_offline_packet.zip.sha256",
    "run_packet_index": "run_packet_index.v0.1.json",
    "counsel_packet_index": "index.json"
  }
}
Decision receipts

Every run ships forwardable artifacts: a human report, a machine triage summary, and a counsel packet index that can be verified offline.

Declared scope

Every run declares exactly what was checked, when, and against which corpus snapshot.

Deterministic outputs

Same inputs produce the same packet structure and hashes, enabling reliable diffs over time.

Audit-ready delivery

Artifacts are designed to be forwarded to counsel and reviewed independently.

90-second review demo

Review a public IP event before you install anything.

The Review Console is the fastest way to understand LocusTrace. Start with the public sample event, inspect evidence and coverage, record one disposition, then open the live console or starter dataset when you want a local packet export tied to the source data hash. The starter dataset path is `review-console/examples/review_console_mrna_litigation.v0.1.json`.

REVIEW

Review the sample event

Start with the public 90-second sample. No account, runtime, or CLI is required.

OPEN

Open the live console

Use the hosted console or `review-console.html` from the starter ZIP for local datasets.

DECIDE

Record one disposition

Mark the event watched, escalated, dismissed, or flagged.

EXPORT

Export the evidence trail

Download the reviewer packet or disposition ledger for verification.

Dataset status confirms the loaded JSON contract.
Items and Data hash stay visible during review.
Ledger hash changes only when the reviewer disposition changes.
Reviewer packet export remains tied to the source data hash.
Evaluation offer

Send one public watch area. Get a packet back.

The fastest path is a bounded 2-week evaluation: one watch area, one baseline packet, one drift update, and a direct review of whether the evidence model fits your workflow.

  • >>One baseline LocusTrace Evidence Packet
  • >>One drift-update packet against the same declared watch area
  • >>A manifest with hashes, source references, and artifact inventory
  • >>A technical debrief on whether the workflow is worth buying
Why it matters

Patent-sequence drift monitoring: why it matters

Patent risk is not static.

The moment a new application publishes, a claim is amended, or a family expands into a new jurisdiction, your exposure changes. A one-time patent search or periodic review is not defensible when risk evolves continuously.

Most patent tools optimize for exploration — dashboards, filters, similarity scores. They do not optimize for proof.

When a question becomes: “What changed, when did it change, and can you prove it?” Most tools cannot answer.

That is why continuous patent monitoring needs receipts, not screenshots.

Omnis LocusTrace is built specifically for that problem.

What makes LocusTrace different

Omnis LocusTrace treats patent monitoring as an evidence pipeline, not a UI.

Every monitoring run produces deterministic, replayable artifacts that answer three questions unambiguously:

  • • What changed
  • • Why it changed
  • • What exactly was examined

Those answers are captured as structured outputs, not screenshots or transient dashboards.

Inspection-ready by design

Each run emits a complete, self-contained record:

  • • A run manifest describing inputs, parameters, corpus coverage, and outputs
  • • A run digest summarizing material changes in human-readable form
  • • Hash-addressed evidence artifacts for every detected delta
  • • Adapter provenance, including adapter identity, contract version, and configuration hash
  • • Pinned corpus snapshots with explicit coverage and content hashes

Nothing is implicit. Nothing is inferred later.

Deterministic and replayable

Omnis LocusTrace is deterministic by construction:

  • • Same inputs produce the same outputs
  • • Outputs are sorted, canonicalized, and hashed
  • • Multi-sequence watchlists execute in a stable, documented order
  • • Run identifiers and artifact paths are predictable and reproducible

This means results can be replayed, compared, and audited long after the fact.

Offline verification, no trust required

LocusTrace outputs can be verified offline, without:

  • • access to source code
  • • access to private repositories
  • • access to the registry
  • • trust in the operator

A verifier only needs the run directory. Using the built-in verification tooling, reviewers can independently confirm:

  • • schema validity
  • • artifact integrity (hash checks)
  • • corpus consistency
  • • adapter contract compatibility
  • • run completeness

This is critical for legal review, procurement, and air-gapped environments.

Signed execution, public verification

Omnis LocusTrace ships as public standalone runtime bundles and a signed container image.

  • • Standalone bundles are published on the public mirror for Linux x64 and macOS
  • • Signatures are keyless, issued via GitHub Actions OIDC
  • • The image and its signatures are publicly verifiable
  • • Verification does not require credentials

This provides a clear chain of custody from source to execution to output.

Built for audit, not dashboards

Omnis LocusTrace intentionally avoids hiding complexity behind visual layers.

Instead, it produces artifacts that stand on their own:

  • • suitable for internal review
  • • suitable for external counsel
  • • suitable for regulatory or litigation contexts

Dashboards can be built later. Evidence cannot.

In short

Omnis LocusTrace exists because:

  • • Patent risk is continuous
  • • Proof matters more than presentation
  • • Verification should not depend on trust

If you need inspection-ready, replayable evidence of patent change over time, this is the system designed for that job.

Protocol

Not just software. A protocol.

Omnis LocusTrace is the reference implementation of a protocol for machine-checkable patent-sequence evidence. The artifact format is designed to outlive the company that created it.

Independent verification

Multiple implementations (Node.js, Python) prove the specification is unambiguous. Same inputs produce identical hashes across languages.

Frozen semantics

Claim classes A-E are frozen until v0.2. Normative boundaries are documented. No swamp monster schemas.

  • • Class A: Publication ingestion
  • • Class C: Overlap/diff
  • • Class E: Legal monitor
Citable artifacts

Bundles have canonical citation formats that survive Word/PDF/email chains. Lawyers can cite pc:ep:v0.1:sha256:3f8c... in briefs.

Honest status
✅ Proved
  • • Another team implemented verifier (Python)
  • • Another machine reproduces hash (CI determinism)
⚠️ In Progress
  • • External field test (awaiting first dry run)
  • • Live workflow citation (templates ready)

The infrastructure is complete enough. Now we test the social reality of the object.

Context

Why it exists

Most teams treat sequence-to-patent checks as a one-time snapshot.

LocusTrace answers the question other tools do not
  • What changed since the last run
  • Why does it matter
Why snapshots fail

That fails because the patent landscape moves while your program is still moving.

Acquisition paths

Start from the workflow buyers search for

Each path leads to the same free trial, but the framing matches the program, counsel workflow, or platform integration need.

Execution model

What it does

You give us sequences tied to a program as watchlists. We monitor public patent sequence corpora over time and detect material deltas.

Material deltas we detect
  • New publications in known families
  • Family expansions across jurisdictions
  • Similarity band changes
  • Coverage shifts
  • Region overlap changes
Evidence packets, not screenshots

Every delta comes with an Evidence Packet: scope, provenance, and similarity traces in canonical JSON so you can rerun and diff results over time.

New workflow

Engineered enzyme and IVT patent analysis

Omnis LocusTrace now includes an enzyme-platform lane for mRNA manufacturing disclosures: WIPO sequence-listing ingest, patent-text provenance, T7-based candidate ranking, and typed comparisons in a single deterministic report.

Operator flow
  • Start from a signed sequence-listing evidence pack or a `WO...` publication via `compile-wipo`.
  • Use the pinned T7 RNAP preset by default, so the common operator path is zero-config on the reference side.
  • Generate one deterministic report with candidate ranking, patent-text provenance, raw artifact references, and typed comparisons.

This workflow is designed for platform IP review: what was disclosed, which sequence is the plausible polymerase candidate, and which patent-text sections support the finding.

PCT to report

Compile directly from a WIPO publication or from an existing signed evidence pack with the same report contract.

Patent-text provenance

Abstract, claims, and description each carry source, extraction method, confidence tier, hash, and raw-artifact path.

Polymerase candidate ranking

Disclosed proteins are ranked against canonical T7 RNAP using offset-tolerant motif and N-terminal signals before full review.

Typed comparisons

DNA and RNA disclosures are not treated as protein hits; mismatches are surfaced explicitly instead of producing invalid alignments.

Deliverables

What you get every run

Three deliverables every run: a human digest, a machine summary, and Evidence Packets you can forward and audit.

report.md

Human-readable posture report with explicit interpretation guardrails.

alerts_summary

Machine-readable triage summary (plus deterministic NDJSON/CSV views for routing).

Counsel packet

Single deliverable with hashes + schemas + evidence refs; independently verifiable offline.

Differentiators

What makes this different

Built as an evidence contract: declared scope, deltas, determinism, and receipts you can forward.

Monitoring, not search

One search is a demo. A watchlist is a budget line item.

Platform-enzyme lane

The same evidence model now handles polymerase and IVT patent disclosures, not just sequence watchlists.

Deltas, not hit lists

Family level dedup and cooldown suppression to avoid alert spam.

Declared scope

Every run declares what was checked and what was not. No ambiguity.

Reproducible by design

Deterministic pipeline, stable ordering, canonical JSON, no system clock dependence.

Security friendly delivery

Signed standalone bundles plus a signed container image, with receipts, SBOM, and provenance.

What this is not

To be explicit:

  • Does not produce legal opinions
  • Does not claim full recall
  • Does not replace counsel review

It produces evidence and change receipts so teams can decide sooner.

Who uses it

Sequence-heavy programs where drift is expensive:

  • Therapeutic proteins
  • mRNA manufacturing and IVT enzyme teams
  • Gene therapy and cell therapy constructs
  • Diagnostic assay sequences
  • IP counsel reviewing polymerase and process-patent disclosures
  • Platform teams generating many variants

Free 2-week trial

Start with 1 to 3 LocusTrace Watchlists, ship deterministic Evidence Packets across a 2-week trial, then decide whether to buy the production key.

  • 1 to 3 watchlists
  • Issued evaluation key
  • Weekly cadence
  • Threshold tuning and one triage session per week
  • Deliver report.md, alerts_summary, and counsel packet artifacts
  • Production-key decision at the end of the trial
Inputs required from customer (keeps scope tight)
  • Upstream similarity results in BLAST TSV, MMseqs TSV, or CSV
  • Metadata map
  • Corpus manifest declaring source, jurisdictions, date range
Assurance

Trust and verification

This is intentionally built as a contract artifact, not a black box.

Verification & Audit Guide →
Verify results offline, without source code or registry access.
Contract artifacts
  • Contract anchor: Diff Engine Spec v0.1 at commit 2bdaed1
  • Reference implementation: v0.1.1 at commit 63a2e0d
  • Public runtime bundles: patentchecker-linux-x64.tar.gz and patentchecker-macos-universal.tar.gz (v0.1.52)
  • Signed runner image: ghcr.io/omniscoder/patentchecker@sha256:1b3e02309d2...
Addressing the obvious question: is it proven

Omnis LocusTrace is new. The right way to evaluate it is not testimonials. It is artifacts.

  • Download and unpack a public standalone runtime bundle
  • Download demo_offline_packet.zip + demo_offline_packet.zip.sha256
  • Verify ZIP sha256
  • Run packet lint with the shipped CLI
  • Optionally verify the signed runner image and attestations

If the receipts hold up, the tool holds up.

Pricing

Free trial first, production key second. Pricing is scoped from watchlists, cadence, and deployment mode instead of monthly self-serve tiers.

FAQ

LocusTrace FAQ

Short answers for teams evaluating watchlists, cadence, evidence, and offline verification.

What is continuous patent monitoring?

Continuous patent monitoring is a watchlist workflow that checks for new publications, amendments, and family expansions on a fixed cadence so teams can react to material change instead of re-running one-off searches.

How is continuous patent monitoring different from a patent search?

A patent search is a snapshot. Continuous patent monitoring is a repeatable process: declared scope, stable inputs, and comparable outputs over time so you can answer what changed and when.

What does Omnis LocusTrace output each monitoring run?

A human report (report.md), a machine-readable alert summary (alerts_summary), and a counsel packet index that enumerates every artifact with hashes for offline verification.

Can Omnis LocusTrace analyze mRNA manufacturing and enzyme patents?

Yes. Omnis LocusTrace includes an enzyme-platform workflow for polymerase and IVT disclosures: WIPO sequence-listing ingest, patent-text provenance, T7-based candidate ranking, and typed comparisons that explicitly flag DNA or RNA versus protein mismatches.

Why do deterministic Evidence Packets matter?

Because counsel and R&D need receipts they can forward, audit, and verify offline. Deterministic artifacts reduce ambiguity about what was checked and make historical comparisons defensible.

Is Omnis LocusTrace legal advice or a freedom-to-operate determination?

No. Omnis LocusTrace is informational evidence infrastructure that produces monitoring artifacts; it does not provide legal opinions, infringement determinations, infringement analysis, or FTO conclusions.

Proof

Verify a demo packet (offline)

Download a single deliverable, verify hashes + schemas + evidence refs offline, and keep the receipts.

Public demo catalog

Live drift pages, delivery-risk scoring, and signed proof bundles now sit under one LocusTrace demos hub so buyers can inspect the whole surface instead of a single artifact.

PTAB appeal evidence bundle

A signed, replay-verifiable bundle capturing a PTAB reversal decision: structured metadata, decision PDF, portal captures, provenance, and a replay verifier.

Delivery IP risk scorecard

A deterministic viral-vector blocking readout for a Prime-style CNS program: capsid overlap, licensing dependency, manufacturing sensitivity, and exclusivity zones in one forwardable report.

NAVSEA proof bundle

A signed, deterministic proof bundle for a naval anti-biofouling coating use case: structured inputs, evidence artifacts, risk events, and a cross-platform replay verifier.

Verifiable risk timeline, not vibes

A historical in silico simulation derived from public-like patent artifacts and synthetic protected assets. No license required to verify.

example_platform_risk_bundle — Cumulative risk score (0–100) Events: 13 Window: 2016-01-01..2021-01-01 Predicate mix: manual 0, alerts_delta 8, claim_signal 5 0 25 50 75 100 2016 2017 2018 2019 2020 2021 Risk posture changes from monitor → engage Similarity band increased t… Claims broaden toward anyRNA encoding spike-like ant… Claims expandnucleoside-modified RNA lan… Similarity band increased t… Claims broaden toward anyRNA encoding spike-like ant… Claims expand nucleoside-mo… Risk score capped at 100 (policy threshold:escalation required) Total risk score: 100 experiment_id: example_platform_risk_bundle.v1 contract_hash: sha256:4b78ff6129907f3e201fe18849a1591e4bfefd9039f969780de02dbdac588b42 schema_digest: sha256:382125195a4532513eff5ea13956c6d8ba0e95b07ab63553f9d5a5c221f0147c
Verify offline: unzip the bundle and run the verifier command below.
Downloads (verify-only)

Two reference bundles: an escalation scenario and a low-signal baseline. Verification is license-free; new analyses are license-gated.

# Assumes the public standalone runtime is unpacked as ./patentchecker.
unzip -q example_platform_risk_bundle.v1.zip -d bundle_dir

# Verify (no license required).
RUNS="$(ls -d bundle_dir/run/run_* 2>/dev/null | LC_ALL=C sort | paste -sd, -)"
./patentchecker/patentchecker risk verify \
  --ontology bundle_dir/example_platform_risk_ontology.v0.1.json \
  --timeline bundle_dir/example_platform_risk_timeline.v0.1.json \
  --runs "$RUNS" \
  --strict --format json