Botrotation — AI Agent Bot Rentals

Reviewed 1d ago.

• tools: 6 tools scanned
• metadata: scanned

No findings.

Tools (6)

• discover_botrotation

  Read Botrotation's public service overview before evaluating a rental. Returns capabilities, dated research counts, selection/evidence limits and payments.methods, the full supported asset/network list with invoice identifiers. Use subscription_plans for price/slots, evaluate_bots for current bot metrics, and agent_onboarding for registration and purchase instructions. No arguments, authentication or payment required; this read does not create an invoice or account. Upstream failure returns an MCP tool error, not an empty offer.

• subscription_plans

  Read current subscription pricing and access allowances before requesting an invoice. Returns items with plan_id, USDC reference price, period_seconds, concurrent_slots, rotation/renewal policy, plus plan_hash and collection_enabled. A held collection state prevents new payments. Use discover_botrotation for payment routes and agent_onboarding to purchase through the delegated REST/SDK workflow. No arguments, authentication or payment required; this read neither subscribes nor reserves slots. Upstream failure returns an MCP tool error.

• evaluate_bots

  Inspect current VFT dashboard evidence and rental admission before selecting bots. Call with {} for board IDs, metric definitions and freshness (items is empty), then select a board or exact bot_id; both filters intersect. Returns complete dashboard rows, ranks, sizing, curves and rental.catalog_admitted. Follow next_offset until null. Unknown bots return an empty items list. Stale/unavailable evidence is labelled and must not be treated as current eligibility. Use bot_history for dated results of the offered cohort, including withdrawn versions. Free, unauthenticated read; grants no strategy licence or execution authority. Invalid arguments/upstream failure return an MCP tool error.

• bot_history

  Read permanent offered-bot histories to assess results after selection, including losses and withdrawals. Default returns version summaries with baseline dates, source status, latest observations and coverage gaps; page using next_offset. events=true returns dated observations/changes with integrity hashes; page using next_after_id. A null cursor ends either sequence; unknown exact filters return empty items. Legacy first-offer dates may be unknown, and simulated VFT observations are sampled, not full trade ledgers. Check freshness and gaps. Use evaluate_bots for current dashboard rankings. Free, unauthenticated read; does not change admission or customer deployments. Invalid arguments/upstream failure return an MCP tool error.

• rental_terms

  Read the complete current Terms of Service before owner-signed delegation. Returns Markdown content, its lowercase hex SHA-256, public company identity and offering status. Use sha256 unchanged for signed acceptance; do not hash reformatted text or substitute subscription_plans' plan_hash. Use agent_onboarding for the acceptance/registration workflow. No arguments, authentication or payment required; reading does not accept terms, register an agent or create a subscription. Upstream failure returns an MCP tool error.

• agent_onboarding

  Read the integration workflow after reviewing bots, subscription_plans and rental_terms. Returns api_base_url, OpenAPI URL, ordered workflow.steps for owner-signed delegation, registration, invoicing/payment and customer execution, plus public SDK/engine package, checksum and guide links. Resolve relative links against api_base_url. This is shared documentation, not account-specific status or an onboarding action; no arguments, authentication or payment required. It does not register, sign, purchase, download or deploy anything. Perform those steps through the returned REST/SDK interfaces with the owner's bounded authority. Upstream failure returns an MCP tool error, never a successful workflow containing an error.