Open Grant Proposal: TRELYAN Filecoin Archival Layer — Post-Quantum Inscription Cold Storage

[brandonjsellam-Releone]

Open Grant Proposal: TRELYAN Filecoin Archival Layer

Project Name: TRELYAN Filecoin Archival Layer — long-term cold-storage tier for post-quantum inscribed artifacts

Proposal Category: Storage

Individual or Entity Name: TRELYAN Foundation (in formation, Zug, Switzerland) — submitted by founder Brandon Joseph Sellam pending Stiftung incorporation. Counsel of record: LEXR Zürich. The Foundation will execute the grant agreement post-Stiftung-registration (target Q3 2026); pre-registration, Brandon Joseph Sellam will sign as the natural-person founder if Filecoin Foundation accepts that bridge.

Proposer: @masterkaizer

Project Repo(s):

• Public site + whitepaper: https://trelyan.foundation
• Repository for grant deliverables (will be created on grant approval): github.com/trelyan-foundation/filecoin-archive (Apache-2.0)

(Optional) Filecoin ecosystem affiliations: None at submission. TRELYAN uses IPFS (Pinata + web3.storage hot-pinning) as the artifact persistence layer today; this proposal funds the Filecoin-native cold-storage tier that sits below the hot-pin layer.

(Optional) Technical Sponsor: None at submission. We would welcome a Filecoin Foundation team member as technical sponsor; the natural fit would be someone on the FVM / storage-deal automation side.

Do you agree to open source all work you do on behalf of this RFP under the MIT/Apache-2 dual-license? Yes. All code under Apache-2.0; all documentation under CC-BY 4.0.

---

Project Description

TRELYAN is a post-quantum inscription protocol on Algorand mainnet. The protocol issues 1,024 numbered non-fungible Vault Cells, each carrying the irrevocable right to inscribe one immutable artifact (≤ 4 KB payload) under a Falcon-1024 lattice signature, verified at consensus by Algorand's native falcon_verify opcode (Algorand mainnet, AVM v12, deployed November 2025). Each inscribed artifact is meant to outlive the protocol's founder; the artifact's persistence has to outlive any single hot-pinning provider. Public site: https://trelyan.foundation.

This grant proposal funds the Filecoin-native long-term cold-storage layer that sits below TRELYAN's IPFS hot-pinning tier. The thesis: a post-quantum signed artifact that lives only on hot-pin IPFS infrastructure has the operational fragility of a CDN; a post-quantum signed artifact additionally archived to Filecoin storage deals has the redundancy that matches the protocol's permanence claim.

Value

This grant funds the Filecoin-specific operational layer that distinguishes a post-quantum inscription protocol with hot-pinned-only artifact storage (operationally fragile) from one with cold-stored archival redundancy (operationally permanent). The latter is what TRELYAN's value proposition requires.

For the broader Filecoin ecosystem, the project produces:

• A clean public example of a non-DataDAO Filecoin-archival pattern (most current Filecoin storage-deal documentation assumes a DataDAO context; this is a per-artifact / per-inscription pattern that hasn't been productized for the inscription-protocol use case)
• A worked integration between Algorand-substrate smart contracts and Filecoin storage deals — useful for the broader cross-chain conversation about how non-Filecoin L1 protocols can credibly use Filecoin storage
• Apache-2.0 reusable tooling that any inscription protocol on any L1 substrate can use to add Filecoin archival

Deliverables (Apache-2.0 code, CC-BY 4.0 documentation)

1. trelyan-filecoin-archive — Python CLI + library for creating Filecoin storage deals for inscribed Cell artifacts. Inputs: IPFS CID from TRELYAN's hot-pin layer. Outputs: Filecoin deal CID + retrieval-verification report.

2. Retrieval-verification integration — Helper that verifies a Filecoin-archived artifact matches the on-chain inscribed artifact_hash invariant in TRELYAN's smart contract. Critical because the inscription contract enforces a 32-byte hash binding at consensus; the archival layer must preserve byte-for-byte fidelity.

3. FVM-side verification stub — A small Filecoin Virtual Machine (FVM) contract that allows third parties to query, on-chain on Filecoin, whether a given Cell artifact has been archived to Filecoin. Useful for due-diligence reviewers, FINMA pre-rulings, and any third party verifying TRELYAN's permanence claim. Deployed to Calibration testnet during grant period.

4. Cross-substrate adapter scaffolding — Generalises the patterns to ML-DSA (FIPS 204) and SLH-DSA (FIPS 205) signatures, not just Falcon-1024 (FIPS 206 track). The Filecoin archival pattern is then reusable across the entire post-quantum signature landscape, not just Algorand's Falcon path.

5. Public documentation site — Static-site documentation under CC-BY 4.0 explaining how to layer Filecoin cold storage on top of any IPFS-pinned inscription protocol. Generic patterns + worked example.

6. Operational runbook — Storage-deal renewal cadence, retrieval-test schedule, monitoring dashboard for archive health. Required for TRELYAN's Stiftungsrat to discharge its archival-permanence fiduciary duty; useful for any project with long-term-archival commitments.

7. 18-month maintenance + security-patch commitment post-release.

Development Roadmap (6-month delivery from grant disbursement)

Month	Deliverable	Output
1	trelyan-filecoin-archive CLI + library v1 (full, not v0.1)	Apache-2.0 GitHub repo
2	Retrieval-verification integration + adversarial test harness	Tested against TRELYAN's contract on TestNet
3	FVM verification stub deployed to Calibration testnet	Public deployment + ABI
4	Cross-substrate adapter scaffolding (ML-DSA + SLH-DSA)	Apache-2.0 release
5	Documentation site v1 + operational runbook	Public site under CC-BY 4.0
6	Open-source release + community announcement	Forum post on Filecoin Discourse + Algorand forum

Total Budget Requested

USD $50,000 (the Open Grants ceiling)

Budget breakdown:

#	Line item	Hours	Rate	Cost
1	Engineering: trelyan-filecoin-archive CLI + library — full v1 production-ready	160	$120/hr	$19,200
2	Retrieval-verification integration + adversarial test harness	60	$120/hr	$7,200
3	FVM verification stub (deployed to Calibration testnet)	50	$130/hr	$6,500
4	Cross-substrate adapter scaffolding (ML-DSA + SLH-DSA generalisation)	40	$130/hr	$5,200
5	Documentation site + integration guide + worked example (CC-BY 4.0)	50	$100/hr	$5,000
6	18-month maintenance + security-patch + community-support budget	50	$120/hr	$6,000
7	Storage-deal testing costs + Calibration testnet runs + retrieval-test fleet	—	—	$1,000
8	Reserve for emergent issues (API breakage, FVM upgrade, dependency churn, license-compatibility questions)	—	—	(rounded down)
	Total			$50,000

Maintenance and Upgrade Plans

18 months of explicit maintenance budget post-release. At month 17, we publicly document whether continued maintenance will be supported by the TRELYAN Foundation's operating budget (post-Cell-pre-sale, Q4 2027+) or transition to a community-maintainer model. Transparency on this is itself an output.

Team

Brandon Joseph Sellam, Founder. Decade of daily KYC/AML / source-of-funds discipline in physical-commodities brokerage under FINMA-supervised counterparty regimes; prior Swiss financial-industry experience at Messodie Zug; engineer-trained (Columbia 2019, full-stack). Native or near-native EN/FR/IT. Authored TRELYAN's whitepaper, modern algopy/PuyaPy/AVM v12 inscription contract, internal threat-model pass (T1-T28).

Cryptographer advisory bench — outreach in motion to Thomas Prest (PQShield), Zvika Brakerski (Weizmann), Vadim Lyubashevsky (IBM Research Zurich), Léo Ducas (CWI), Pierre-Alain Fouque (Univ. Rennes), all emailed 27 May 2026. One of these is expected to convert into an advisor role over the next 90 days. Vadim Lyubashevsky is the designated primary co-fiduciary candidate (Swiss-resident, FIPS-204 credentials).

Smart-contract auditor — RFQ in motion to Runtime Verification (lead), Trail of Bits, OpenZeppelin, CertiK, Halborn, Coinspect, Ulam Labs (all 7 sent 27 May 2026). Fixed-price quote expected in 2-3 weeks.

AI Intelligence Council — three-seat advisory configuration (Anthropic Claude Opus, IBM watsonx Llama 3.3 70B, Mistral Large) drawn from structurally independent training lineages. Used for drafting + diligence; does not hold keys or sign transactions.

Team Background

Founder personal capital invested in the project: ZERO (by signed and witnessed Founder Commitment Letter Sections 1-3, executed 27 April + 13 May 2026). The funding stack is exclusively:
(a) non-dilutive grants (NLnet NGI0 Commons Fund €50,000 application 2026-06-2da submitted 27 May 2026; Algorand Foundation Accelerator queued for next cohort opening; this Filecoin Foundation grant; other cycle applications as appropriate)
(b) F&F bridge round USD 25,000 – 100,000 (in preparation)
(c) Cell pre-sale Q4 2027 post-audit, up to ~USD 3.69M aggregate

Relevant Use Cases, Stories, Research, or Implementations of Storage / Retrieval / FVM

TRELYAN's archival use case sits at the intersection of three Filecoin product surfaces:

1. Storage deals — per-artifact, per-inscription storage commitments. Each of the 1,024 Cells produces (at most) one inscribed artifact; the archival pattern is one-storage-deal-per-Cell, with renewal automation managing the long-term commitment. This is a NEW pattern compared to DataDAO patterns (which are typically batch-DataSet-oriented) and a NEW pattern compared to single-large-dataset patterns (Filecoin Slingshot, etc.).

2. Retrieval-verification — once a Cell is inscribed, the on-chain Algorand contract enforces a 32-byte hash binding at consensus. The Filecoin archive must preserve byte-for-byte fidelity; the retrieval-verification helper confirms this on each retrieval test. This is a clean public example of how to verify a non-Filecoin L1 substrate's hash invariant against Filecoin-archived data.

3. FVM — the verification stub allows third parties to query, on-chain on Filecoin, whether a given artifact has been archived. Useful for due-diligence reviewers, regulatory pre-rulings (FINMA), and any third party verifying a long-term-archival permanence claim.

Additional Information

• Public site: https://trelyan.foundation
• Whitepaper: linked from public site
• Contact: brandonjsellam@releone.io
• Foundation status: Swiss Stiftung in formation (Zug); counsel of record LEXR Zürich; TRELYAN trademark filed at IGE Zürich under application 08601/2026 (Nice classes 9 / 36 / 42 — protected scope explicitly names NIST FIPS 206 in Class 9 + Class 42 specifications). IGE accelerated examination underway; decision expected ~19 June 2026.
• Parallel non-dilutive applications: NLnet NGI0 Commons Fund (€50,000, application 2026-06-2da submitted 27 May 2026 for the open-source toolkit layer SEPARATE from this Filecoin scope); Algorand Foundation Accelerator (queued for next cohort).

Cross-Chain / Compatible Networks

This grant is Algorand → Filecoin specifically (TRELYAN's substrate is Algorand mainnet; Filecoin provides the archival tier). The patterns are designed to generalise to any L1 substrate that pins ARC-3 / ERC-721-style inscribed artifacts: the templates are deliberately substrate-agnostic at the integration layer. The cross-substrate adapter scaffolding (deliverable #4 above) makes this generalisation a first-class output, not an afterthought.

---

Founder commitment letter signed pages, threat-model pass T1-T28, audit-firm RFQ template, NLnet submission record, IGE trademark filing receipt all available on request via brandonjsellam@releone.io for due-diligence verification.