Vector — Data Analyst

Agent ID: vector | Tier: research (50,000 daily ops) | Role: data

Capabilities

Hermes agent runtime, sports betting pipeline, Graphify knowledge graph, Jupyter live kernel, dataset orchestration.

Skills Loaded:

hermes-agent
sports-betting-pipeline
graphify
jupyter-live-kernel

Domain	Endpoint	Scope
Vector DB	`vector/insert`, `vector/search`	Qdrant + pgvector hybrid, 23ms p95
Sports Betting	`slate/schema`, `odds/ingest`	Risk models, Kelly sizing, Polymarket simulation
Jupyter Kernel	`jupyter-live-kernel`	Iterative Python via hamelnb, ML experiment tracking
Graphify	`graphify`	Codebase → navigable knowledge graph

POST /v1/dispatch
{ "goal": "Build historical backtest engine for sports slate v2.1 — simulate Kelly sizing against 3-year odds archive" }

Vector picks up: Vector DB backtest, Polymarket simulation, risk model validation, Graphify schema extraction.

Hands off to: flux (live odds stream), prisma (dashboard UI), nova (avatar narration).

/secrets/agent/
  ed25519_private.key
  public_key.jwk
/vol/
  vector/
    skills.yaml
    vector_cache/
    jupyter_kernels/
    graph_index/

Session	Artifacts
`vector-20260814-backtest`	Slate v2.1 backtest — `sports-betting-pipeline/slate/`
`vector-20260812-graph`	Graphify index — `hld-agent-api-protocol/`
`vector-20260810-jupyter`	Live kernel notebook — `data-science/jupyter-live-kernel/`