Capabilities / Programs

Holder quests

Verify completion before a reward is proposed.

Implementation boundary

BSC adaptation

Strict QUEST budgets/requirements and full-holder D1 qualification/Merkle allocation run through actual API and model dispatcher; real local activation/claims/refunds exist. Model prepare is unsigned and paid-source fixture is explicitly separated. Remaining: Automatic signer grant for community activation/commit/claims is absent; requirements support fixed reviewed kinds, not arbitrary natural-language verified tasks. Original private quest classifier/eligibility filters are unknown.

Current awake evidence

actual local http or sqlite evm subset verified

Production verification: not complete. Registered model tools: create_quest · close_quest · inspect_community. Registration does not mean enabled or actually run.

Original typed tools

create_quest

Published limits / Solana reference

{
  "max_program_available_pct": 100
}

Original capability specification ↗

These are source-reported limits. BSC values and venue limits require calibration.

Back to registry