Raxx · internal docs

internal · gated ↑ index

NDA framework — decisions, alternatives, and sourcing

ATTORNEY REVIEW REQUIRED BEFORE USE. This document is research-only, produced by a non-lawyer. It surfaces decisions, tradeoffs, and primary-source citations so the user can walk into a consult with a contract attorney (SaaS / technology-transactions background, licensed in the chosen governing state) and get to judgment calls quickly. None of the positions taken here are legal advice. The companion template (nda-template.md) is NOT a finished contract — it's the strawman the attorney will review, modify, and bless.

Status: research-only. Last updated: 2026-04-22. Verify all citations and cost figures at decision time.


TL;DR


1. What kind of NDA — one-way vs mutual

Decision: one-way for v1

Epic #161 scopes v1 as one-way (Disclosing Party = Raxx; Receiving Party = signer). This is appropriate because the three v1 use-cases — alpha access gating, investor conversations (early-stage, pre-term-sheet), and contractor onboarding — all have a one-directional information flow: Raxx shares proprietary material; the counterparty is receiving it in exchange for consideration (access, evaluation opportunity, payment).

When mutual becomes necessary (v2 trigger)

A mutual NDA becomes necessary when the counterparty will disclose their own confidential information in a substantive way. Typical triggers:

For v1, none of these are expected. If one arises, software-architect + product-manager should scope a v2 mutual template as a separate engagement and not force-fit v1.

Asymmetry notes (tradeoff to flag to attorney)

A one-way NDA where the Disclosing Party (Raxx) grants injunctive relief and attorneys' fees to itself but not to the signer is common and defensible, but it is signer-unfriendly. Sophisticated counterparties (VCs, corporate-dev teams) will push back and demand mutuality at least on remedies and governing-law convenience. The template as drafted uses two-way attorneys' fees ("prevailing Party") as a middle position — attorney should advise whether to soften further or accept pushback on a case-by-case basis.


2. Carve-out philosophy — what counts as Confidential Information

Two ends of the spectrum

Template position: loose + reasonable-person standard

The drafted §1.1 uses the loose approach with a reasonable-person backstop ("that a reasonable person would understand to be confidential or proprietary given the nature of the information and the circumstances of disclosure"). This is the dominant pattern in tech / SaaS NDAs because:

  1. Pre-launch startups rarely have the discipline to mark every email, Figma frame, or Zoom conversation.
  2. Courts are generally comfortable with the reasonable-person test for breach analysis.
  3. Receiving parties have less leverage to negotiate for strict marking when the context is alpha-access or early investor exposure.

Standard four exclusions (tracks industry consensus)

Section 1.2 uses the standard four carve-outs:

The "compelled by law / subpoena" category is placed in §2.5 (permitted disclosure with notice) rather than as an exclusion, which is the more modern pattern — compelled disclosure is not a carve-out from what counts as confidential, it's a permitted disclosure channel.

Industry framing

Common confidentiality terms can range between 2, 3 and 5 years, with two to five years commonly accepted as a reasonable period for safeguarding most confidential information. Five years is a common length in nondisclosure agreements that involve business negotiations and product submissions, although some companies prefer three or two years. — EveryNDA, "Duration Clauses in Non-Disclosure Agreements" [2]


3. Trade-secret-tail clause (DTSA + state UTSA)

Rather than a flat term, the template uses a dual-track approach in §5.2:

This avoids the "perpetual NDA = unenforceable" problem in states that disfavor indefinite restrictions, while preserving indefinite protection for genuine trade secrets.

A recommended approach is a standard 3-5 year term with a carve-out stating that information qualifying as a trade secret remains protected "for so long as such information constitutes a trade secret under applicable law." This approach avoids problems with perpetual NDAs while protecting genuine trade secrets indefinitely. — EveryNDA, synthesized industry position [2]


4. Term length — 2 / 3 / 5 years

Quick-compare

Term Posture Defensible? Tradeoffs
2 years Signer-friendly Yes, common in vendor/agency NDAs Too short for investor-targeted material if roadmap / financial projections look beyond 2yr horizon; trade-secret tail still protects core IP
3 years Balanced default Yes, dominant for SaaS / tech Matches typical roadmap visibility (12–24 months out + buffer); signer less likely to pushback than 5yr
5 years Disclosing-Party-friendly Yes, common in manufacturing / licensing Strong pushback from sophisticated signers; may deter some alpha-user signings; tech information often stale before the 5yr clock ends

Industry context

Short-term NDAs (1–2 years) are common in vendor and agency partnerships, mid-term NDAs (3–5 years) are used in tech collaborations and licensing talks. In technology and software, where things change rapidly, NDAs tend to be for one to three years because tomorrow's breakthrough concept may become obsolete in a matter of moments. In contrast, in pharmaceuticals or manufacturing, where research and product development can take decades, NDAs can be for 10 years or more. — mydock365 / EveryNDA / Certinal synthesis [1][2][3]

Recommendation to carry into consult

3 years + trade-secret tail is the defensible middle for Raxx's profile:

Attorney to confirm the specific value for [TERM_YEARS]. No strong reason to deviate from 3 unless attorney identifies a jurisdiction-specific concern.


5. Governing law — PA vs DE (or other)

The conventional picks

Pennsylvania:

Delaware:

Delaware is great for corporate law governing company structure and shareholder rights, but those laws don't govern IT transactions, which are governed by state contract law and federal IP law, where Delaware has no special advantage. — Tech Contracts, "Don't Choose Delaware Law Unless You're in Delaware" [4]

Less-common alternatives to flag

Recommendation to carry into consult

Pennsylvania, once the PA LLC is formed. Pre-formation (if the portal launches before [ENTITY_NAME] exists), attorney should advise on whether to:

[GOVERNING_COUNTY] will follow from the entity's registered-office county.


Federal baseline: ESIGN Act (15 U.S.C. § 7001)

Enacted in 2000. Core rule in § 7001(a):

A signature, contract, or other record relating to such transaction may not be denied legal effect, validity, or enforceability solely because it is in electronic form ... [or] solely because an electronic signature or electronic record was used in its formation. — 15 U.S.C. § 7001(a), via Cornell LII [5]

Applies: nationwide, to transactions "in or affecting interstate or foreign commerce" — which Raxx's web-based NDA flow is.

Does not apply to: wills, codicils, testamentary trusts, divorce / adoption / family-law matters, court orders, utility cancellation notices, health-insurance / life-insurance cancellations, product-recall notices, and documents required by law to accompany hazardous material transport. None of these affect Raxx NDA scope.

ESIGN § 7001(c) imposes specific consumer-consent disclosures when a law otherwise requires that the record be provided in writing to a consumer. For commercial-context NDAs (between a business and an alpha-user/investor/contractor), § 7001(c) is usually not directly triggered because no separate law requires the NDA itself to be in writing to a consumer — the NDA is the first instrument creating the relationship. But:

The consumer-consent requirements per § 7001(c)(1) include:

(A) the consumer ... has affirmatively consented to such use and has not withdrawn such consent; (B) the consumer, prior to consenting, is provided with a clear and conspicuous statement ... (C) the consumer ... consents electronically, or confirms his or her consent electronically, in a manner that reasonably demonstrates that the consumer can access information in the electronic form that will be used ... — 15 U.S.C. § 7001(c)(1), via Cornell LII [5]

State layer: UETA

Enacted in 49 states + DC + USVI. Illinois adopted UETA in June 2021, replacing its prior Electronic Commerce Security Act [6]. Washington adopted UETA in 2020 (SB 6028). New York is the lone remaining holdout and relies instead on the Electronic Signatures and Records Act (ESRA) [7][8].

UETA's key operative sections for NDA enforceability:

Practical ESIGN+UETA elements the portal must implement

For the signed NDA to hold up:

  1. Intent to sign — clear "I Agree" / "Sign" action separated from mere site navigation.
  2. Consent to electronic records — affirmative checkbox or equivalent unambiguous action before the signer is presented with the NDA.
  3. Attribution — email verification, IP, user-agent, timestamp captured and stored.
  4. Record retention — tamper-evident storage (SHA-256 content hash at sign-time); retrievable on signer request.
  5. Ability to retrieve — signer can download their own signed copy at any time.
  6. Consent withdrawal mechanism — for future records (not the signed NDA, which remains binding); see ESIGN § 7001(c)(1)(B)(ii).

All six are operationalized in esign-compliance-checklist.md.

New York exception (if signer is NY-based)

New York's ESRA is substantively similar to UETA but uses different statutory language. The ESIGN Act preempts contrary state law except where UETA (or a substantially-similar state statute) has been adopted, so ESIGN applies as the federal floor in NY regardless. Practically: NY signers are fine under ESIGN.

Despite not adopting UETA, New York ... [is] nevertheless governed by the E-SIGN Act pursuant to federal law. — ZapSign state-level compilation of eSignature laws [7]

Attorney to confirm if Raxx plans to gate significant NY-based signer volume and whether NY-specific consent-flow language is worth adding.


7. DTSA whistleblower-immunity notice (18 U.S.C. § 1833(b))

The DTSA, enacted in 2016, requires employers to include a notice of whistleblower immunity in contracts with employees, contractors, or consultants that restrict use/disclosure of trade-secret information [10][11].

An employer that fails to provide the mandated notice may not be awarded exemplary damages or attorney fees in trade secret litigation against an individual to whom such notice was not provided. — Katten / Troutman Pepper Locke synthesis of 18 U.S.C. § 1833(b)(3)(C) [10][12]

Does it apply to Raxx's NDA use-cases?


8. Remedies — injunctive, damages, attorneys' fees, arbitration

Injunctive relief

Template §6.1 acknowledges irreparable harm and entitles Disclosing Party to seek injunctive relief. This acknowledgment lowers the bar for obtaining a TRO/preliminary injunction — in many states (including DE), stipulations to irreparable harm are generally sufficient to establish that element of the injunction standard [13].

"Without posting a bond" language: the template drafts it in favor of Disclosing Party. Courts vary on whether this waiver is enforceable; some require a nominal bond regardless. Attorney to advise.

Attorneys' fees

Template §6.3 uses "prevailing Party" (two-way). Alternatives:

Two-way is the balanced pick. Attorney should confirm PA / [GOVERNING_STATE] enforceability of prevailing-party clauses generally.

Arbitration is not included in the template. Reasoning:

Attorney to confirm. If Raxx ever needs to swap in arbitration (e.g., for investor-facing v2 NDA where the investor prefers AAA commercial arbitration), it's a surgical replacement of §8.2 + addition of an arbitration clause.

Liquidated damages — not included

Liquidated damages (a fixed dollar amount per breach) are common in some industries but tricky:

Attorney to confirm.


9. Dispute-resolution flow (informal)

Not in the template, but worth flagging as an optional add the attorney may want:

Recommendation: do NOT include cure-period or meet-and-confer as a condition precedent to injunctive relief. Confidentiality breaches are time-sensitive; any precondition on court access weakens the primary remedy. If included at all, carve out injunctive relief from the precondition.


10. Open questions for the attorney consult (aggregated)

These roll up all questions from this doc + nda-template.md. Intended to be taken as a single agenda into the #151 consult.

Engagement + scope

  1. Can you handle NDA template review + revision as part of the entity-formation / trademark package (from docs/business/questions-for-attorney.md), or is this a separate scope / flat fee?
  2. Fee for reviewing, marking up, and signing off this template as Raxx's production version?
  3. Turnaround time?

Structural decisions

  1. Term length: 3-year headline + trade-secret tail — agree, or prefer 2 / 5?
  2. Governing state: PA (once LLC is formed) vs. DE — agree with PA, or reasons to prefer DE?
  3. Pre-formation naming: if portal launches before [ENTITY_NAME] exists, use Kris personally with post-formation assignment, or delay portal until formation lands?

Drafting specifics

  1. Carve-out philosophy: loose + reasonable-person standard (current §1.1) vs. marked-only. Which matches standard attorney practice for this context?
  2. Non-use §2.2: non-solicitation subclause (§2.2(c)) — keep standalone, move to a separate agreement, or strike? (Question sharpens if any alpha users are CA-resident.)
  3. Section 5.2 trade-secret tail — confirm formulation matches PA UTSA / DTSA language.
  4. Section 6.1 "no bond" language — keep, soften, or strike?
  5. Section 6.3 attorneys' fees — two-way ("prevailing Party") vs. one-way in favor of Disclosing Party?
  6. Section 7 DTSA notice — confirm inclusion is appropriate for all three v1 use-cases.
  7. Section 8.3 jury-trial waiver — keep, strike, condition on jurisdiction?
  8. Section 10 ESIGN language — sufficient for commercial-context NDA, or should § 7001(c)(1)(A)-(D) consumer-consent disclosures be added to the portal consent flow as belt-and-suspenders?

Use-case coverage

  1. Should v1 bifurcate into distinct templates for (a) alpha users, (b) investors, (c) contractors? Or is one template with the same language fine for all three?
  2. v2 mutual-NDA trigger — confirm that v1 one-way only is appropriate and flag any v1 use-case that secretly needs mutuality.
  3. Contractor use-case — is a plain NDA sufficient, or does it need to be paired with a contractor IP-assignment agreement (tying into docs/business/questions-for-attorney.md §E)?

Portal mechanics (for architect handoff)

  1. Section 10.4 record retention — is "commercially reasonable time" acceptable, or should we commit to a specific SLA (e.g., "within 5 business days")?
  2. Section 10.3 attribution — is IP + UA + email + timestamp sufficient attribution, or do we need additional factors (e.g., SMS 2FA, device fingerprint) to meet UETA § 9 for higher-value signers (investors)?
  3. Record retention period — template is silent; epic #161 calls for NDA term + 6 years. Confirm 6-year buffer is appropriate for PA (4-year § 5525 SoL + 2-year buffer).

11. Sources

Primary sources:

Secondary / explainers:


ATTORNEY REVIEW REQUIRED BEFORE USE. Carry this document and nda-template.md into the #151 consult with a contract attorney (SaaS / technology-transactions background, licensed in [GOVERNING_STATE]). The #10 open-questions list is the consult agenda.