Synthesize a protocol brief from the research notes

Ranking

Hard filters first (including MCP tool policy), then weighted score. Policy denials are eligibility failures, not score penalties.

Events

21

Policy evaluated

0

Policy denied

0

1. TASK_SUBMITTED2026-09-28T22:58:10.144Zschema 1.0

   Synthesize a protocol brief from the research notes

2. RANKED2026-09-28T22:58:10.161Zschema 1.0

   Selected protocol-synthesis (agent:research.synthesis.v1@1.0.0) score 99.9 using balanced role synthesize. Best match id "synthesis"; outcomeTrust 0.76 n=50 weight 6 base 95.3 effective 99.9; audit passed.

   agent:research.synthesis.v1@1.0.0

3. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.164Zschema 1.0

   No capability, skill, tag, or keyword matched the task requirements

   agent:orchestrator.core.v1@0.9.0

4. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.167Zschema 1.0

   No capability, skill, tag, or keyword matched the task requirements

   agent:pricing.engine.v1@0.8.2

5. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.169Zschema 1.0

   No capability, skill, tag, or keyword matched the task requirements

   agent:contract.escrow.v1@0.7.0

6. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.172Zschema 1.0

   No capability, skill, tag, or keyword matched the task requirements

   agent:robot.bay.arm.v1@1.0.0

7. HARD_FILTER_REJECTEDinactive2026-09-28T22:58:10.175Zschema 1.0

   Status is maintenance — only active agents are eligible

   agent:compute.market.v1@0.6.0

8. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.177Zschema 1.0

   No capability, skill, tag, or keyword matched the task requirements

   agent:devtools.docsreader.v1@1.0.0

9. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.180Zschema 1.0

   No capability, skill, tag, or keyword matched the task requirements

   agent:research.gpuaas.v1@1.0.0

10. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.183Zschema 1.0

    No capability, skill, tag, or keyword matched the task requirements

    agent:devtools.issuewright.v1@1.0.0

11. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.185Zschema 1.0

    No capability, skill, tag, or keyword matched the task requirements

    agent:research.literature.v1@1.1.0

12. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.188Zschema 1.0

    No capability, skill, tag, or keyword matched the task requirements

    agent:memory.vault.v1@1.1.0

13. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.191Zschema 1.0

    No capability, skill, tag, or keyword matched the task requirements

    agent:devtools.mystery.v1@1.0.0

14. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.193Zschema 1.0

    No capability, skill, tag, or keyword matched the task requirements

    agent:privacy.router.v1@0.5.1

15. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.196Zschema 1.0

    No capability, skill, tag, or keyword matched the task requirements

    agent:research.provenance.v1@1.0.0

16. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.199Zschema 1.0

    No capability, skill, tag, or keyword matched the task requirements

    agent:devtools.repoadmin.v1@1.0.0

17. HARD_FILTER_REJECTEDno_capability2026-09-28T22:58:10.201Zschema 1.0

    No capability, skill, tag, or keyword matched the task requirements

    agent:trust.score.v1@0.4.0

18. SELECTED2026-09-28T22:58:10.204Zschema 1.0

    Winner protocol-synthesis (agent:research.synthesis.v1@1.0.0) effective 99.9.

    agent:research.synthesis.v1@1.0.0

19. ROUTED2026-09-28T22:58:10.212Zschema 1.0

    Routed to https://agents.example/protocol-synthesis via a2a.

    agent:research.synthesis.v1

20. RESULT_RECEIVED2026-09-28T22:58:10.222Zschema 1.0

    Result success from agent:research.synthesis.v1.

    agent:research.synthesis.v1

21. TASK_DONEfinal2026-09-28T22:58:10.228Zschema 1.0

    Task success by agent:research.synthesis.v1.

    agent:research.synthesis.v1

Result

Standardized agent result. Stored by taskId; retries keep prior attempts.

{
  "kind": "research.synthesis",
  "brief": "Spot GPU slices cut expected spend but can be reclaimed, so long jobs pay a preemption tax. Burst or exploratory jobs usually fit spot; training runs with costly checkpoints usually fit reserved. Reserved GPU slices trade a commitment cost for a lower chance of interruption and more predictable start times. The tradeoff is interruption risk versus commitment, not a vendor catalog. Sources: doc:gpuaas-spot-reserved — Spot versus reserved GPU tradeoffs.",
  "query": "Spot vs reserved GPU capacity?",
  "reason": "strong_topical_match",
  "privacy": "restricted",
  "sources": [
    "doc:gpuaas-spot-reserved"
  ],
  "summary": "Synthesized 4 focused claims from 1 source.",
  "corpusScope": "public",
  "upstreamStep": "research",
  "evidenceStatus": "sufficient"
}