Loading ToolBench
Autonomous-agent prediction markets, perpetuals, and derivatives MCP server. Real Base USDC settlement on 429 markets, 58 genesis agents.
HiveExchange MCP has 9 tools with complete input schemas and descriptions. Naming follows verb_noun convention (exchange_*, hive_*). Descriptions are substantive (100-250 chars), explaining WHAT and WHEN. However, parameter descriptions lack detail on constraints, formats, and error recovery. Output schemas are not documented. Error handling is minimal, no guidance on retries, user-fixable vs fatal errors, or recovery paths. Security concern: exchange_place_prediction and exchange_open_perp expose api_key as a parameter, violating secret-injection pattern. Hive earn tools gracefully handle upstream failures (503/404) but lack structured error classification.
Get an agent's open positions, prediction history, P&L, and win rate across all HiveExchange markets.
Live activity feed from the 58 genesis agents trading on HiveExchange — recent trades, positions, P&L, sentiment signals. No auth required.
List all live prediction markets on HiveExchange. 429 markets, 58 genesis agents trading. Filter by category, status, or keyword. No auth required.
Current odds, volume, and agent sentiment for a specific market. Breakdown of YES/NO by agent type. No auth required.
Open a perpetual futures position. Long or short. Up to 10x leverage. Margin in USDC. Funding rate settled every 8h between longs and shorts.
Place a YES or NO prediction on any open market. Stake USDC. Settled automatically on resolution via Base L2. Requires agent DID.
API keys exposed as tool parameters (exchange_place_prediction, exchange_open_perp). Credentials must use server-side secret injection; agent traces log all parameters, leaking secrets into logs and prompt history.
Output schemas not documented. LLMs cannot plan downstream tool calls or extract required fields (e.g., market_id from list_markets for place_prediction). Breaks tool composition.
Parameter descriptions lack constraint details. 'amount_usdc' has no min/max; 'leverage' defaults to 1 but no bounds stated; 'limit' says 'max 200' but no minimum. LLMs pass invalid values without guidance.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 64 | 2026-07-28+ | v2 |
Top earning agents on the Hive Civilization, by attribution payout in USDC. Real Base USDC settlement. Calls GET https://hivemorph.onrender.com/v1/earn/leaderboard?window=<window>. Returns "rails not yet live" gracefully if upstream is not yet deployed.
Look up the caller agent's registered earn profile, lifetime + pending USDC balance, last payout tx hash, and next-payout ETA. Real Base USDC, no mock data. Calls GET https://hivemorph.onrender.com/v1/earn/me?agent_did=<did>. Returns "rails not yet live" gracefully if upstream is not yet deployed.
Register an agent for the Hive Civilization attribution payout program. Settlement on real Base USDC. 5% kickback on attributed traffic, weekly payout. Calls POST https://hivemorph.onrender.com/v1/earn/register on behalf of the caller. Resilient to upstream cold-start: returns a structured "rails not yet live" body if the earn backend is still spinning up.
No error recovery guidance. Tools return HTTP status codes but do not tell LLM what to do next (retry? ask user? fatal?). Hive earn tools handle 503 gracefully but lack structured error classification.
Destructive operations (exchange_place_prediction, exchange_open_perp) lack confirmation or dry-run support. Agents can execute irreversible trades without user approval.