Demand-response for agent fleets: contracted shed, verified settlement, paid interruptibility.
Price the hold, not just the use: demurrage tariff engine for agent fleet resources.
Pre-submission escrow receipts and coverage-aware retention diffs.
Browser-download delivery for agents: captures stranded downloads, delivers with a signed receipt.
Durable ownership of an approved agent PR: signed verdicts, give-up report by cause.
Cross-session fatigue ledger for agents: FIT/CAUTION/GROUNDED dispatch gating.
Compensation registry for agent actions: inverses registered before effects land, with rewind.
Signed dispatch releases: routes legal only within earned diversion reach of verified fallbacks.
Signed receipts + horizon statements proving whether an event-triggered agent actually woke.
Dead-man's switch for agent fleets: tranche-gated succession on multi-channel absence proof.
Signed agent self-forecasts, resolved and scored per cohort, with a self-killing trust weight.
General average for agent fleets: bonded, rating-neutral sacrifice loss-sharing.
Certified local pilot boards for an unfamiliar site, with scoped authority and signed completion.
Sealed capsules and a wake gauntlet for rarely-run agents returning to a changed world.
Query-intent privacy for agents: threshold cloaking with signed exposure receipts.
Salvageability triage for agent fleets in a declared mass-failure mode.
Calibrated Glicko-2 per-dimension agent rating with deviation, from verified receipts only.
Fail-closed per-task-class dispatch matrix with deferral clocks and signed dispatch releases.
Multilateral barter netting for agent services: cycles and chains cleared in kind, zero money.
Checkpoint-paired twin trials for agent fleet changes with blinded verdicts.
Deadline-priced slow lane for deferrable agent work: futures, dispatch-or-credit, signed journal.
Standing minted only from signed settlement receipts, never reviews. Verdicts: ACCEPT to DECLINE.
Cache-bust forensics + TTL-fit verdicts: names the field that busts your prompt cache.
Propagation breaker: egress quarantine + fingerprint outbreak telemetry for fleets.
About AI & Agents MCP servers
This category collects servers that make agents better at being agents. You will find orchestration layers that run multi-step pipelines, routers that hand a task to the right model or sub-agent, evaluation and tracing tools, prompt and skill libraries, and bridges that let one assistant call another. Some wrap commercial model APIs; others are thin coordination layers around open-source frameworks.
Because these servers often sit between your agent and other systems, look closely at where data flows and what is logged. Prefer servers that document their tool contracts clearly, support streaming where relevant, and can run locally when the workload is sensitive. Many are new and move quickly, so check the repository's recent activity before relying on one in production.