Market Trust diligence packet
Will the highest score achieved on Humanity’s Last Exam in 2026 be 60% or higher?
One of four bands, strongest warning on the right
market quality score.
Use only as a review target. Current-card issues must clear before this price is used as evidence.
Market quality
44 / 100
Interval 5-54
Evidence confidence
low
50 / 100
Source health
source gate passed
1 sources | dropped 0
Audit chain
a99191ab58...
row hash verified
Memo language
Convexly rates this polymarket market "Will the highest score achieved on Humanity’s Last Exam in 2026 be 60% or higher?" as Do not cite. Use only as a review target. Current-card issues must clear before this price is used as evidence. Caveat to carry: Current card issue: Not enough pillars on this card are measured yet. This is market diligence, not a trading instruction or proof that the market probability is correct.
Caveat to carry: Current card issue: Not enough pillars on this card are measured yet.
How to use this packet and what changes the verdict
Use this packet when
- Use to explain why this market should stay out of a citation set for now.
- Use as a recovery checklist for source freshness, coherence, depth, or evidence gaps.
Do not use it as
- A trading instruction.
- Proof that the market probability is correct.
- A manipulation accusation or person-level identity claim.
What changes the verdict
- Clear current blocker: Not enough pillars on this card are measured yet.
- Clear current blocker: Some pillars on this card are still gathering evidence.
- Clear current blocker: The rating moves under leave-one-out or weight-perturbation checks.
How the score adds upWhy this verdict
Why this rating
Composite 44 · Do not cite
- Coherencemeasuredweight 25.0
2 full-provenance Coherent Markets Engine signals linked to this condition; 11 price-observation checkpoints recorded.
- · full provenance cme signal linked
- · signal price observations linked
- · no observed checkpoint yet
30+10.7 - Executabilityheuristicweight 20.0
Executability scored as worse-of(spread, depth) from a single orderbook snapshot; the two were merged from the former liquidity + depth pillars to stop double-counting one correlated source.
- · executability worse of spread depth
- · spread: top-of-book spread 1.0c
- · spread: spread is 3% of mid, tight relative to price
- · spread: pp-band-to-cents-band approximation in effect (depth_5pct_usd/2 per side); per-side decomposition pending substrate work
15+4.3 - Resolution reliabilityheuristicweight 20.0
Resolution source and rule text are present but not independently reviewed.
- · resolution metadata present manual review pending
70+20.0 - Participant flowpendingweight 20.0
Participant flow inputs exist but do not meet identity-coverage floors; descriptive only.
- · below publishable sample floor
- · edge score composite descriptive not skill
- · high participant concentration
- · insufficient scored accounts
n/ano contrib - Integrity risk screenpendingweight 10.0
Heavy concentration / wash-pattern signals; treat with caution.
- · top-3 wallets control 100% of notional flow
- · 100% of wallet/side pairs appear repeatedly
- · flow events: 6
- · distinct wallets: 2
55no contrib - Audit completenessmeasuredweight 5.0
3 linked artifacts; 0 missing hashes; 0 stale/failing sources.
- · artifact hashes present
- · source watermarks healthy
100+7.1
How it adds up: composite = Σ(score × weight) / Σ(weight) across measured + heuristic pillars only. Pending and data-rights-blocked pillars are excluded from both numerator and denominator. Total contributing weight = 70.0.
Caveat: 2 pillars pending; the score will recompute as the substrate fills in.
Market identityWhat is being rated
Venue
polymarket
condition 0xcb251bef81c1eaa9...
Category
AI
used for coverage and validation breadth
Market status
active
venue-provided lifecycle state when available
Resolution rule
This market will resolve to "Yes" if any model achieves at least the specified accuracy on Humanity’s Last Exam by December 31, 2026, 11:59 PM ET. Otherwise, this market will resolve to "No". For resolution, “accuracy” refers solely to the value labeled “HLE Accuracy” or a clear equivalent metric if the data’s presentation or terminology is restructured, regardless of the model’s Calibration Error or any other displayed metric. The resolution source will be the official Humanity’s Last Exam leaderboard at https://agi.safe.ai/. If the resolution source becomes unavailable during the listed timeframe, this market will remain open to allow the relevant data to become available again. If the source remains unavailable after the end of the listed timeframe or is otherwise confirmed to be permanently unavailable, official Humanity’s Last Exam results published elsewhere may be used. If no official alternative source is available, this market will resolve to "No".
card_id: mt-v02:polymarket:0xcb251bef81c1eaa9267a4bc8b9811922722d9afc9a6e7f5477b2167884011ba2:2026-08-06
market_id: a7871da5-f8aa-47c5-85ac-a1d29dda4ac2
venue_market_id: 0xcb251bef81c1eaa9267a4bc8b9811922722d9afc9a6e7f5477b2167884011ba2
resolution_source: https://agi.safe.ai/
Audit chainReceipt and verification
Convexly stores the card payload, removes the row_hash field, canonicalizes the remaining JSON, and recomputes the SHA-256 hash. This is the bridge from a readable verdict to a buyer-reviewable audit chain.
method: market_trust_card_v0_2
row_hash: a99191ab58272e26b7c197eac71b4e5537ebdef0557d4ce2ca582260a228c8dd
previous_row_hash: f098cff7a8f4a8ab601546bf965c9c0dba01c870951086d8c9b28a984b2390eb
source_artifact_hashes: 43a375c67ed1b3fe1e73dbb87dde61086f3a1fa0e658bfbc440d82e65cf8e326, 74f4d4ba9daae7ee7fd67434aafcfe9d7cdc803bb42f4f8c693e5ad36ee125e2
Verdict hashImmutable verdict archive
This verdict is archived as a canonical JSON object with a SHA-256 content hash. The hashed payload carries the methodology version (market_trust_card_v0_2), the canary_preview label, and a fixed boundary statement; none of them can be removed without changing the digest.
verdict archive sha256: 4674a44ebcf546dbb306f7fd6bb68549807226b4ba05bf8d67abe67542ef225f
How to verify this hash
- Fetch the canonical archive bytes: https://www.convexly.app/api/market-trust/v0.2/0xcb251bef81c1eaa9267a4bc8b9811922722d9afc9a6e7f5477b2167884011ba2/archive?venue=polymarket&format=canonical
- Recompute the digest locally, for example: curl -s "https://www.convexly.app/api/market-trust/v0.2/0xcb251bef81c1eaa9267a4bc8b9811922722d9afc9a6e7f5477b2167884011ba2/archive?venue=polymarket&format=canonical" | shasum -a 256
- Compare your digest to archive_sha256 in the JSON response at https://www.convexly.app/api/market-trust/v0.2/0xcb251bef81c1eaa9267a4bc8b9811922722d9afc9a6e7f5477b2167884011ba2/archive?venue=polymarket. Matching digests mean the archived verdict content is byte-for-byte unchanged.
- Cross-check card_row_hash against the stored full-card receipt at https://www.convexly.app/api/market-trust/v0.2/0xcb251bef81c1eaa9267a4bc8b9811922722d9afc9a6e7f5477b2167884011ba2/verify?venue=polymarket.
Stored v0.2 cards are frozen scorecards: a re-score inserts a new row with a new card_id and records the prior receipt in previous_row_hash, and the filed calibration design binds Convexly to never mutate frozen card-time rows. The archive is a deterministic serialization of one stored row, so the same row always reproduces the same archive_sha256.
Participant and integrity screenWho shaped the price
Participant-flow screens flagged concentration, unknown-flow, or low-edge-flow conditions for review. This is an integrity-risk screen, not a manipulation accusation and not person-level identity resolution.
Scored flow
67.3%
1 scored accounts
Unknown flow
32.7%
higher unknown flow lowers confidence
Participants
2
public accounts observed
Top flow share
67.3%
concentration diagnostic
Not claiming: wash-trading detection; sybil detection; manipulation-free market.
Market Fitness - pre-trade diligence (experimental)
4 of 6 diagnostics have a derivable observable on this stored card; 2 are insufficient_data. An insufficient_data row means the card lacks that input, not that the check passed.
Depth and slippage at size is a Researcher feature.
The resolution-reliability, manipulation, and basis-risk reads below are open. To see the fill cost at $100 / $1k / $10k and the executability detail, upgrade to Researcher.
Depth durability
heuristic
insufficient data
At-size depth and slippage detail is a Researcher feature. Upgrade at /pricing to see fill cost at size.
Observable: Executability pillar: worse-of(top-of-book spread, modeled price impact) from the latest inspected orderbook snapshot at 2026-08-06T18:45:11.555+00:00
Caveats: derived from a single orderbook snapshot; durability across time is not yet measured on this card; descriptive diagnostic from operator-set scoring that is not calibrated; no interval exists for this value, so none is shown rather than fabricated; gated detail requires researcher
Spread and price impact
heuristic
insufficient data
At-size depth and slippage detail is a Researcher feature. Upgrade at /pricing to see fill cost at size.
Observable: Top-of-book spread (price points) and modeled price impact for a fixed notional, from the latest inspected orderbook snapshot
Caveats: spread cost over a holding period is not modeled; this is a single-snapshot read; descriptive diagnostic from operator-set scoring that is not calibrated; no interval exists for this value, so none is shown rather than fabricated; gated detail requires researcher
Time to resolution
insufficient data
insufficient data
Venue lifecycle status on the card: active. No end date is stored.
Observable: Venue-supplied market end date or close time stored on the card
Missing input: The v0.2 card schema does not store an end date or close time, so time to resolution cannot be derived from this card.
Resolution reliability
heuristic
70 / 100 (descriptive)
Resolution source and rule text are present but not independently reviewed. Resolution rule text is stored on the card.
Observable: Resolution reliability pillar: review of the resolution source (https://agi.safe.ai/) and stored resolution rule text
Caveats: resolved-outcome dispute history is not joined on this card; the read covers resolution source and rule text only; descriptive diagnostic from operator-set scoring that is not calibrated; no interval exists for this value, so none is shown rather than fabricated
Concentration exposure
insufficient data
insufficient data
Participant-flow screens flagged concentration, unknown-flow, or low-edge-flow conditions for review.
Observable: Participant-flow concentration from the card's integrity screen: top account flow share, unknown flow share, scored notional share
Missing input: Participant-flow concentration is not measured on this card (no scored participant rollup or no top-account flow share).
Basis-risk note
descriptive
structural note
This market settles as a binary event contract: it pays a fixed amount on one discrete resolution. Offsetting a continuous exposure with a binary payoff leaves residual basis risk by construction, and binary payoffs will generally fail the ASC 815 80-125% dollar-offset band used for hedge-accounting treatment, so a position that works as an economic offset can still be carried as mark-to-market P&L. This is a structural description of the contract type, not a statement about any party's exposure.
Observable: Contract payoff structure (binary event contract) and the resolution rule text stored on the card
Caveats: structural description of the contract type, not accounting or legal advice
Published deliberately as a component table with no single combined number. Read each component against the observable it derives from; an insufficient_data row means the stored card lacks that input, not that the check passed.
These diagnostics describe observable market-quality measurements for one market. They are not investment advice, not a trading instruction, not a recommendation to take or offset any exposure with this or any market, and not a statement about any person or their positions.
instrument_fitness_diagnostics_v0_1
Source freshness windowIngest status
The current ingest window is clean enough for review. This is freshness reporting, not an uptime SLA.
Latest orderbook
2026-08-06T18:45:11.555Z
most recent orderbook evidence attached to this card
Sources
1
1 healthy | 0 degraded | 0 stale | 0 failing
Dropped events
0
nonzero dropped events block current-language claims
Observed age
4s
oldest observed source watermark
Ingested age
1s
oldest ingest watermark
Validation stateCanary vs v1
Canary pipeline
blocked
Current card has source, evidence, or robustness issues to clear.
Public v1 promotion
4 public v1 validation gates still open.
Canary readiness says the current batch can be reviewed. Public v1 promotion waits for coverage, clean source-health windows, resolved outcomes, and formula-freeze evidence.
Read the promotion evidence contract- Not enough pillars on this card are measured yet.
- Some pillars on this card are still gathering evidence.
- The rating moves under leave-one-out or weight-perturbation checks.
- 14 consecutive clean source-health days have not yet been observed.
- 30 current cards across 3 market categories are not yet reached.
- The scoring formula must be frozen before validation runs.
- At least one resolved-outcome row is still needed in the validation ledger.
Top caveatsWhat travels with the score
- Current card issue: Not enough pillars on this card are measured yet.
- Current card issue: Some pillars on this card are still gathering evidence.
- Current card issue: The rating moves under leave-one-out or weight-perturbation checks.
- 100% of wallet/side pairs appear repeatedly
- artifact hashes present
- below publishable sample floor
Before stable v1Validation requirements
- 14 consecutive clean source-health days have not yet been observed.
- 30 current cards across 3 market categories are not yet reached.
- The scoring formula must be frozen before validation runs.
- At least one resolved-outcome row is still needed in the validation ledger.
External contextNon-scoring provenance
No external baseline family is mapped for this market yet.
No external baseline family is mapped for this market yet.
External context is non-scoring in v0.2. Market Trust still measures market-quality substrate, not absolute truth against outside forecasts.
Full seven-pillar evidence grid
Coherence
weight 25 · measured
30 / 100
2 full-provenance Coherent Markets Engine signals linked to this condition; 11 price-observation checkpoints recorded.
Executability
weight 20 · heuristic
15 / 100
Executability scored as worse-of(spread, depth) from a single orderbook snapshot; the two were merged from the former liquidity + depth pillars to stop double-counting one correlated source.
Resolution reliability
weight 20 · heuristic
70 / 100
Resolution source and rule text are present but not independently reviewed.
Participant flow
weight 20 · pending
pending
Participant flow inputs exist but do not meet identity-coverage floors; descriptive only.
Integrity risk screen
weight 10 · pending
55 / 100
Heavy concentration / wash-pattern signals; treat with caution.
Audit completeness
weight 5 · measured
100 / 100
3 linked artifacts; 0 missing hashes; 0 stale/failing sources.
Coherence blockLinked-market consistency
2 full-provenance Coherent Markets Engine signals linked to this condition; 11 price-observation checkpoints recorded.
Score
30
measured evidence
Source refs
2
linked-market or CME artifacts
full provenance cme signal linked signal price observations linked no observed checkpoint yet
Robustness blockSensitivity and drivers
Sensitivity verdict
moves
leave-one-out and weight perturbation
Leave-one-out
35-53
score range
Weight perturbation
42-45
score range
Drivers
3
largest weighted contributors
- Coherence scores below 60 and needs stronger evidence before the verdict can improve.
- Executability is heuristic, not fully measured.
- Executability scores below 60 and needs stronger evidence before the verdict can improve.
- Resolution reliability is heuristic, not fully measured.
- Participant flow is pending, not fully measured.
- Integrity screen is pending, not fully measured.
Live market integrity read is a Trader feature
Trader ($499/mo) unlocks an on-demand participation-concentration read on any market — HHI of wallet volume shares with a confidence interval on every estimate, plus a round-trip self-churn share, computed live from the current flow tape. Descriptive structure with receipts, not a manipulation call.
Boundary of the packet
This packet is market diligence, not investment advice, not a claim of absolute truth, and not proof that any trader should buy or sell. It answers whether the market price has enough substrate to cite, monitor, discount, or ignore with receipts attached.