HALOWERK govwerk

Endpoints (10 live)

• POST /v1/proposal-summary — Splits caller-supplied proposal text into sentences, scores each sentence by normalized non-stopword frequency, and returns the highest-scoring sentences in source order. It does not infer unstated intent, verify claims, or produce an abstractive interpretation. (0.003 USDC on Base)
• POST /v1/delegate-alignment — Joins caller-supplied vote vectors by proposal ID, encodes against/abstain/for as -1/0/1, and computes weighted cosine similarity plus exact agreement. It describes only supplied overlapping votes and does not infer ideology or future behavior. (0.003 USDC on Base)
• POST /v1/snapshot-verify — Sorts unique scalar key/value entries through canonical JSON serialization and computes SHA-256. It verifies byte-level agreement with an optional claimed digest only; it does not prove block inclusion, timestamp, signer identity, or chain consensus. (0.002 USDC on Base)
• POST /v1/token-gini-index — Sorts caller-supplied non-negative balances and computes the standard finite-sample Gini coefficient plus top-one and top-ten-percent shares. Results cover only supplied accounts and do not identify common ownership, custodians, vesting, delegation, or off-chain claims. (0.002 USDC on Base)
• POST /v1/regulatory-check — Joins scalar facts by field name and evaluates explicit equals, not-equals, numeric bound, membership, contains, and existence rules. It does not retrieve law, determine which rules apply, provide legal advice, or certify regulatory compliance. (0.004 USDC on Base)
• POST /v1/vote-buying-detect — Scores caller-supplied vote/transfer observations using cohort-relative transfer size, transfer recency, and repeated-counterparty count. The output is an investigative triage heuristic, not proof of vote buying, identity linkage, or wrongdoing. (0.004 USDC on Base)
• POST /v1/treasury-runway — Subtracts committed liabilities at month zero, then projects recurring inflow and operating outflow with caller-supplied monthly growth rates. It is a scenario calculation, not a forecast of prices, token liquidity, exchange rates, or future governance decisions. (0.003 USDC on Base)
• POST /v1/sentiment-analysis — Tokenizes caller-supplied text and counts matches against a small declared English/German unigram lexicon. It reports coverage and a normalized polarity score; it does not understand context, sarcasm, negation, policy merit, or public opinion. (0.002 USDC on Base)
• POST /v1/voter-turnout-predict — Applies a fixed, published logistic equation to aggregate caller-supplied features and reports an estimated turnout range. It is an untrained scenario heuristic, not a calibrated election forecast, individual propensity score, or basis for voter targeting. (0.003 USDC on Base)
• POST /v1/quadratic-funding-calc — Caps individual contributions when requested, computes each project’s raw quadratic match as (sum sqrt(contribution))² minus direct contributions, and allocates a fixed matching pool proportionally. It does not detect Sybil identities, move funds, or decide project eligibility. (0.004 USDC on Base)