Earnings comparison correction

The displayed profitability passage does not establish an expectations benchmark for the historical EARNINGS_BEAT classification. That comparison is UNSUPPORTED by the displayed passage. The original event is retained; quotation matching, arithmetic and earnings surprise are separate checks.

source_hash identifies the original stored event record; it is not a digest of the displayed excerpt or the original SEC file. Public event inclusion verification: NOT_AVAILABLE. /evidence/verify.json describes SIGNAL snapshot ledger checks and does not authenticate this event hash. Quote occurrence, semantic support, arithmetic and external provenance are separate checks.

Research guide / Guide de recherche

One evidence trace, end to end

A single real filing — Suncor Energy's Q1-2026 results 6-K — followed through every stage of YUCLAW's evidence pipeline. Every value below is pulled live from the same data the public pages render. Suncor is an evidence-tier name: ingested, classified, displayed — never scored.

Disclaimer — Research & education only. Not investment advice. Event types and grades are research classifications, not buy/sell recommendations.

Step 1

SEC filing — SU 6-K

accession 0001104659-26-055746 · accepted by EDGAR 2026-05-06 00:45:00-06:00

Suncor Energy furnished this 6-K to the SEC. YUCLAW's live poller picked it up from the issuer's EDGAR submissions index and recorded it with a canonical accession number (ingestion status: done). Open the primary document on sec.gov →

Step 2

Exhibit → extracted prose

6-K filings carry the substance in EX-99.x exhibits; the envelope has no item codes

exhibit99 — 16,000 characters ingested

          EX-99.2 3 su-20260505xex99d2.htm EX-99.2 All financial figures are unaudited and presented in Canadian dollars unless noted otherwise. Production volumes are presented on a working-interest basis, before royalties, except for production volumes from Suncor Energy Inc.’s (Suncor or the company) Libya operations, which are presented on an economic basis. Certain financial measures in this document, including adjusted funds from operations, free funds flow, net debt and adjusted operating earnings are not prescribed by Canadian generally accepted accounting principles (GAAP). For a description of these non-GAAP financial measures, see the advisory section of Suncor’s Management’s Discussion and…
        

Clean exhibit prose, exactly as stored in the extraction substrate (yuclaw_v5.swarm_inputs) — this text, not the raw HTML envelope, is what downstream classification reads.

Step 3

Classified evidence event

deterministic schema: type · magnitude · direction · confidence · verified excerpt

SU_20260506_2759fceebeac

EARNINGS_BEAT · magnitude 0.8 · direction +1 · extraction confidence 0.9 · status accepted

Source-checked excerpt (SourceLock): “Generated over $4.0 billion in adjusted funds from operations and $2.9 billion in free funds flow.”

Extractor rationale (yuclaw-llm-70b): The company reported record first quarter upstream production, refining throughput, and refined product sales, indicating a strong earnings beat.

Comparison support: UNSUPPORTED_COMPARISON. The cited passage does not explicitly establish the classified comparison with expectations. Original classification retained as a historical record. Extraction confidence is a model output, not a calibrated probability of correctness.

The excerpt is SourceLock-checked against the source text (including documented normalization and similarity fallbacks). This checks quotation support; it does not independently establish an earnings surprise. Event types are research classifications from a locked vocabulary, not recommendations.

Step 4

Evidence grade (coverage depth)

the same member record the Canada Resources page renders — pulled live, not copied

On the Canada Resources page, SUNCOR ENERGY INC currently carries evidence grade A (events + prose evidence) — derived deterministically from coverage depth: 25 filings ingested, 17 with exhibit/MD&A prose, 21 accepted events. Grades describe how much evidence is on file, never desirability. Insider substrate: Outside current evidence scope — SEDI substrate not currently ingested.

Step 5

C6 posture (risk channel)

specialist-swarm reading of the same filing — evidence display, never a score

The v5 specialist swarm read this filing's prose and produced a risk-channel posture — level high / flag normal (aggregation count2, insider gate false). Posture is display-layer evidence context for the Canada page — it never enters composite scoring for evidence-tier names.

Flagged drivers (counts and classifications, no judgments):

Data inconsistency
Declining operating cash flow
FX volatility on USD debt
Inventory valuation volatility
Non-GAAP metric reliance
Non-GAAP reliance
Rising operating expenses
currency/FX

Per-agent contribution: bull: medium · bear: high · skeptic: medium · earningsquality: medium · macro: medium · grounding-verified quotes only (verifier: v5/swarm/grounding.py)

Use YUCLAW in your research

1 · Verify the record

pip install yuclaw then yuclaw replay-lab.
No install: tools/replay_lab.py (stdlib only) against the published bundle.
Exit 0 = core tables reproduced and bundled ledger consistency checked; external provenance is separate. How to report a replication →

2 · Inspect one evidence trace

One real Suncor 6-K, end to end:
filing → exhibit → extracted prose → event type → grade → C6 posture.
Open the trace → · example evidence memo (Suncor) →

3 · Cite a research lens

Every evidence packet ships a ready citation snippet
(version, data-through, build date, source commit).
Get the citation →

Data through 2026-10-14 (last completed U.S. trading day); generated at 2026-10-15 02:45 UTC.

YUCLAW evidence trace · SU 6-K 0001104659-26-055746 · rebuilt 2026-10-15 02:45 UTC · YuClawLab · research & education only