Set the brief, decide how bids will be scored, and choose your committee. You will review everything before publishing.
Both of these are published before bidding and cannot be changed afterwards. Bidders read them, and the models score against them.
What the work is, and what gets delivered.
Escrowed in full. A bid above this is ineligible.
Give each criterion a share of the total score. The weights must add up to 100%. Price is calculated by code. The text criteria are scored by 2 models and checked by code.
Anchors come from the template. They are published with the rubric and are what a model is told each score means.
Anchors come from the template. They are published with the rubric and are what a model is told each score means.
Weights total 100.00%
Bidding closes 24 hours after the tender is created. At least 5 minutes.
Returned unless a bid cannot be decrypted or names another account.
The committee awards the tender and is the only group that can read the full report. Register your own key with a wallet signature, or add another member using their account and public key. A public key must be derived for this tender, not a different one.
No member registered yet. At least one is required.
Ask the member for a committee public key derived for tender 4. Do not paste a wallet private key.
The brief, scoring rules and committee become fixed when you create the tender. Check them now.
These are the exact canonical JSON bytes the contract will store and hash.
{"aggregation":{"disagreement_threshold":3,"evidence_clamp_score":1,"evidence_min_chars":12,"evidence_required_from_score":2,"min_valid_outputs":2,"qualitative":"lower_median"},"award_policy":{"acceptance_window_hours":72,"approvals_required":1,"deviation_requires_rationale":true},"budget_cap_minor":"15000000000","criteria":[{"id":"price","type":"price_lowest_ratio","weight_bp":4000},{"anchors":{"0":"No methodology given.","1":"Generic statements only, nothing tied to this scope.","2":"Some specific steps but major gaps, such as no testing or no threat model.","3":"Clear plan covering manual review and testing, few specifics on NEAR async risks.","4":"Specific plan covering manual review, tests or fuzzing, NEAR cross-contract and callback risks, and reporting.","5":"As 4, plus concrete deliverables, a timeline per phase, and a retest of fixes."},"id":"methodology","question":"How credible and specific is the proposed review methodology for this scope?","source_fields":["methodology"],"type":"qualitative","weight_bp":3500},{"anchors":{"0":"No experience given.","1":"Unverifiable claims only.","2":"General smart-contract experience, nothing NEAR-specific or checkable.","3":"Some NEAR or Rust contract experience, partly checkable.","4":"Several relevant NEAR reviews with checkable public reports.","5":"As 4, plus directly comparable work such as DAO treasury or callback-heavy contracts."},"id":"track_record","question":"How relevant and checkable is the vendor's prior experience for this scope?","source_fields":["track_record"],"type":"qualitative","weight_bp":2500}],"currency":{"asset":"nep141:usdc.sealedchoice.testnet","decimals":6,"symbol":"tUSDC"},"eligibility":{"require_delivery_lte_max":true,"require_price_lte_cap":true,"required_fields":["price_minor","delivery_days","methodology","track_record"]},"max_delivery_days":30,"models":["Qwen/Qwen3.6-35B-A3B-FP8","z-ai/glm-5.3-flash"],"scope":"Manual security review of about 1,200 lines of Rust (near-sdk 5.x): a treasury module with cross-contract withdrawals, allowances and callbacks. Deliverable: written report with severity ratings and one retest of fixes.","tie_break":["total_desc","price_asc","index_asc"],"title":"Security review of a DAO treasury module (near-sdk, Rust)","v":1}