How does DEPLOY's 9-tier source-quality rubric classify verification evidence?

The 9-tier source-quality rubric

DEPLOY's source-quality rubric operates a 9-tier structural classification of verification evidence. The rubric refines the 4-tier source-quality framework from DEPLOY's how-deploy-verifies methodology editorial by distinguishing primary-source subtypes that operate at structurally distinct verification depths within the primary tier. Each primary-source category verifies against different document categories with different jurisdiction + scope + audit-trail properties. Same primary tier; different verification substrates.

The 9 tiers in descending source-quality order:

1. Primary-SEC-filing (10-K + 10-Q + 8-K + S-1 + proxy statements)
2. Primary-FDA-database (510(k) clearance letters + De Novo decisions + premarket approvals + IDE pivotal trial records + Warning Letters + product code records)
3. Primary-government-record (FOIA responses + Army contract awards + USPTO records + Federal Register + agency program documentation + FCC filings + IRS Form 990)
4. Primary-company-IR (audited financial statements + official funding-round press releases + IR-confirmed customer announcements + investor presentations + manufacturer-official bare domains)
5. Primary-academic-peer-reviewed (named-author + journal-venue + verifiable-methodology peer-reviewed publications)
6. Secondary-established-publication (Reuters + Bloomberg + WSJ + FT + AP + established journalistic outlets with named bylines + editorial standards)
7. Secondary-industry-publication (Robot Report + MedTech Dive + FierceBiotech + IEEE Spectrum + industry-trade publications with sector expertise)
8. Aggregator (Wikipedia + SEO content sites + blog aggregators)
9. Unverified (self-report at unestablished outlets + social media claims + reseller marketing without confirmation)

The 9-tier classification operates as the framework substrate for per-claim verification at the source-quality depth where each claim actually resolves.

---

Why 9 tiers, not 4

The 4-tier framework from the methodology editorial canonical reference distinguishes primary + secondary-established + secondary-industry + aggregator categories. The 4-tier framework is sufficient for cohort-level cluster framing where the primary-vs-secondary distinction is the load-bearing verification axis. The 9-tier refinement matters at the per-claim verification layer where distinct primary-source subtypes operate at structurally distinct verification depths.

SEC filings verify financial + corporate-state claims at the audit-trail depth that regulatory disclosure operates. 10-K filings carry annual fiscal-year financial state + executive compensation + corporate-state operations + risk factors with sworn certifications + auditor attestation. 8-K filings carry material-event disclosure within 4 business days of event occurrence. S-1 filings carry registration-statement scope for public offerings. SEC primary-source verification is load-bearing for public-company financial + corporate-state framings.

FDA databases verify regulatory + clearance claims at the device-classification + clearance-scope + post-market-surveillance depth that FDA disclosure operates. 510(k) clearance letters carry product code + intended use + substantial equivalence framings. De Novo decisions carry novel-pathway clearance with regulatory-classification reasoning. Premarket Approvals carry high-risk-device clearance with clinical-evidence trail. IDE records carry pivotal-trial documentation. Warning Letters carry FDA enforcement-action disclosure with specific cited claims. FDA primary-source verification is load-bearing for medical-device + regulatory-scope framings.

Government records verify program + contract + jurisdiction-specific claims at the agency-specific document categories that operate outside SEC + FDA scope. Army contract awards verify defense-procurement at named-contract depth. FOIA responses verify documents not publicly-disclosed at the agency-distribution layer. USPTO records verify patent + trademark + trade-secret framings. Federal Register verifies regulatory-promulgation timing + scope. Government primary-source verification is load-bearing for defense-procurement + non-SEC-non-FDA regulatory + program-specific framings.

Company IR verifies operational + commercial-deployment claims at the management-stated depth that operates within corporate-disclosure scope. Audited financial statements carry IR-confirmed financial state. Official funding-round press releases carry IR-confirmed funding announcements with lead-investor + round-structure detail. IR-confirmed customer announcements carry deployment-attribution detail. Manufacturer-official bare domains (the company's own primary-domain product pages) carry product-specification + feature-claim detail. Per Agent A's taxonomy refinement, manufacturer-official bare domains operate at primary-tier verification depth distinct from secondary-publication coverage of the same product. Company IR primary-source verification is load-bearing for operational + commercial-deployment + product-specification framings.

Academic peer-reviewed verifies methodology + research-claim framings at the journal-venue + peer-review-process verification depth. Named-author + journal-venue + verifiable-methodology peer-reviewed publications carry editorial-process attestation + peer-review verification + replication-capable methodology disclosure. Academic primary-source verification is load-bearing for research-claim + methodology + capability-evaluation framings.

The 5 primary-source subtypes operate at structurally distinct verification substrates. The 4-tier framework collapses them into a single "primary" tier; the 9-tier framework distinguishes them at the per-claim verification layer.

---

The secondary + aggregator + unverified tiers

The 4 non-primary tiers in the 9-tier rubric operate at structurally distinct verification depths from primary-source verification:

Secondary-established publications (Reuters + Bloomberg + WSJ + FT + AP + established journalistic outlets with named bylines + editorial standards) carry substantial verification weight when primary sources aren't directly accessible. The verification depth depends on the outlet's editorial standards + corrections practices + sourcing transparency. When secondary-established sources cite verifiable primary-source attribution, the verification depth approaches but doesn't reach primary-source verification.

Secondary-industry publications (Robot Report + MedTech Dive + FierceBiotech + IEEE Spectrum + industry-trade publications with sector expertise) carry moderate verification weight. Industry-trade coverage adds sector expertise + context that general-audience secondary-established outlets don't carry, but operates at less editorial scrutiny than secondary-established sources. The verification depth varies by publication + by specific reporter + by sourcing transparency.

Aggregator content (Wikipedia + SEO content sites + blog aggregators) carries context-only weight, never primary. Wikipedia entries carry crowd-sourced verification at the Wikipedia editorial-standards depth, which is structurally distinct from journalistic-outlet verification + primary-source verification. SEO content sites operate at content-production-for-search-engines verification depth, which is structurally distinct from claim-verification depth. Blog aggregators carry single-author + variable-editorial-process verification depth. When aggregator framing conflicts with primary sources, primary sources win at every granularity.

Unverified content (self-report at unestablished outlets + social media claims + reseller marketing without confirmation + anonymous tips without verification) carries no verification weight; it's flagged as such or excluded.

The 4 non-primary tiers operate at descending verification weight from secondary-established through unverified. Each tier carries its own verification depth at the source-quality layer where the claim resolves.

---

The per-claim source-quality discipline

The 9-tier rubric provides the source-quality classification substrate; the per-claim discipline determines how claims get verified against the rubric.

Source-quality operates per-claim, not per-source. The same source can carry verified factual claims and unverified allegation framings simultaneously. The framework reads each claim at its own source-quality depth rather than accepting or rejecting sources wholesale.

The canonical worked example: the Red Cat Black Widow SRR contract verification. The Kerrisdale Capital short-seller report (March 2025) on Red Cat operates as a single source that carries claims at structurally distinct verification depths:

• FOIA-anchored figures verified at primary-source depth: the $12.9M LRIP obligated funding figure and the 5,880-systems Army objective quantity resolve at FOIA primary-source verification (tier 3: primary-government-record) regardless of the reporting source's commercial position (short-seller).
• Allegation framings at lower verification posture: management-framing critiques + deployment-posture characterizations + commercial-trajectory assertions operate at single-source verification depth (closer to secondary-industry or unverified tier depending on sourcing transparency).

The framework reads each claim at its own source-quality depth. The short-seller report carries primary-source-anchored FOIA figures + lower-verification-posture allegation framings simultaneously; the framework doesn't reject the FOIA figures because of the source's commercial position, and doesn't accept the allegation framings because the FOIA figures verify at primary-source depth.

Per-source verification would either accept the entire short-seller report at single-source depth or reject the entire report because of the commercial position. Per-claim verification reads the FOIA-anchored figures at tier 3 primary-source verification + the allegation framings at lower-tier verification depths. Same source, multiple verification depths, per the per-claim source-quality discipline.

---

Worked examples per tier

The 9-tier rubric operates with structurally distinct worked examples per tier in DEPLOY's corpus:

Tier 1 (primary-SEC-filing): Intuitive Surgical's 11,395 da Vinci systems + 1,041 Ion systems per SEC 10-Q verifies the canonical installed-base reference at primary-SEC-filing depth. Stryker headquartered in Portage MI per SEC 10-K verifies corporate-state location identification at SEC primary-source depth.

Tier 2 (primary-FDA-database): Happy Ring dual FDA 510(k) clearances K240236 + K242224 verifies the canonical FDA-cleared ring positioning at primary-FDA-database depth. Whoop BPI FDA Warning Letter July 14, 2025 verifies the active regulatory dispute at FDA enforcement-document depth.

Tier 3 (primary-government-record): Red Cat Black Widow $12.9M LRIP obligated per Army FOIA verifies the defense-procurement contract value at FOIA primary-source depth.

Tier 4 (primary-company-IR): Whoop $10.1B Series G March 2026 lead Collaborative Fund verifies the funding-round attribution at IR-confirmed primary-source depth (CORRECTED from earlier aggregator-quoted QIA/Mubadala per Agent A re-verify; the IR-confirmation distinguished the verified lead investor from the aggregator-conflated alternatives).

Tier 5 (primary-academic-peer-reviewed): Physical Intelligence pi0 + pi05_base Apache-2.0 open licensing + published research operates at primary-academic-peer-reviewed depth for the open-source model architecture + methodology verification. Cap-flag on the Happy Ring 97% PSG agreement company-reported: the 97% figure operates at company-reported tier 4 verification (without peer-reviewed citation per Agent A), distinguishing it from the dual FDA 510(k) clearances at primary-FDA-database tier 2 depth.

Tier 6 (secondary-established-publication): when primary-source attribution isn't directly accessible, secondary-established sources (Reuters + Bloomberg + WSJ + FT + AP) carry substantial verification weight. The verification depth depends on the outlet's editorial standards + corrections practices + primary-source attribution transparency.

Tier 7 (secondary-industry-publication): industry-trade coverage (Robot Report + MedTech Dive + FierceBiotech + IEEE Spectrum) carries moderate verification weight + sector expertise depth. The verification depth varies by publication + specific reporter + sourcing transparency.

Tier 8 (aggregator): aggregator framings frequently surface conflations + flat-narrative-framings at aggregator-tier verification depth. The corrections documented across DEPLOY's framework-in-action narratives all operate as catches of aggregator-tier framings against primary-source verification: Stryker Kalamazoo vs Portage; PI $11B vs $5.6B; Covariant acquired-by-Intrinsic vs continuing-but-diminished; Figure 03 BMW Spartanburg vs Figure 02; 1X Redwood humanoid hardware vs brain entity; Monogram Doug Unis CEO vs CMO/founder.

Tier 9 (unverified): self-report at unestablished outlets + social media claims + reseller marketing without confirmation. The verification posture is honest-not-verified rather than treating the claim as carrying verification weight.

The worked examples per tier demonstrate the rubric operating at editorial-anchor depth across the DEPLOY corpus.

---

Aggregator-drift patterns per tier

Each tier attracts structurally distinct aggregator-drift patterns:

Primary-tier drift patterns: primary-source-anchored claims sometimes get aggregator-paraphrased at lower verification depth as the citation chain extends. The original SEC 10-K or FDA 510(k) primary-source claim verifies at tier 1 or tier 2; secondary-publication citation of the primary source operates at tier 6 or tier 7; aggregator paraphrase of the secondary citation operates at tier 8. The verification chain extension can introduce drift at each citation step; per-claim verification against the original primary source resolves the chain at primary-source depth.

Secondary-tier drift patterns: secondary-established publications occasionally publish framings that overstate primary-source verification when the underlying primary-source attribution is incomplete. Secondary-industry publications occasionally carry sector-expertise framings that mix primary-source-anchored claims + sector-judgment framings; the framework reads each claim at its own source-quality depth.

Aggregator-tier drift patterns: aggregator content frequently produces single-narrative framings that collapse multi-fact verification states into single claims (Figure 03 deployed at BMW Spartanburg + Leipzig; PI valued at $11B; Covariant acquired by Intrinsic; Stryker headquartered in Kalamazoo). The framework discipline catches aggregator-tier framings at the primary-source-verification layer; the corrected attribution surfaces honestly at the per-claim depth.

Unverified-tier drift patterns: unverified-tier claims occasionally enter the citation chain through cross-publication propagation. A social-media claim cited in a blog aggregator becomes "reported" content that secondary publications occasionally cite without primary-source verification. Per-claim verification against the original source resolves the chain at the actual verification depth.

---

Why the rubric matters

The 9-tier source-quality rubric matters for the same reason the verified-vs-claimed framework matters: institutional partners audit DEPLOY's framework discipline at the operational-practice layer. Insurance underwriting + data sharing permissions + cross-device interoperability + standards body validation operate at the per-claim verification depth where the source-quality classification actually resolves.

Per-claim source-quality verification produces sharper attributions than per-source classification. The Red Cat short-seller report carries FOIA-anchored figures at tier 3 + allegation framings at lower tiers; per-claim verification reads each at its source-quality depth. The Whoop $10.1B Series G framing carries IR-confirmed lead-investor attribution at tier 4 + earlier aggregator-quoted alternatives at tier 8; per-claim verification distinguishes the IR-confirmed from the aggregator-conflated. The 9-tier rubric provides the framework substrate for per-claim verification at the source-quality depth where each claim actually resolves.

The framework discipline reading source-quality at the per-claim depth is the foundation. The 9-tier rubric refines the substrate; the per-claim discipline applies the substrate to claim-by-claim verification. Together they produce the verification authority that institutional partners audit at the operational-practice layer.

For the methodology editorial canonical reference where the 4-tier source-quality framework is described, see how DEPLOY verifies. For the canonical Red Cat Black Widow worked example demonstrating per-claim source-quality discipline (FOIA-anchored figures + allegation framings within the same short-seller report), see How DEPLOY corrected the Red Cat Black Widow contract value. For the parallel methodology essays operationalizing structurally distinct framework discipline layers, see Captive vs third-party brain providers + Autonomy-boundary classification + Verified-vs-claimed at within-entity granularity.