360 Secure SkillHub
360 安全技能中心

pontx-massive

面向 Massive 市场数据的集成型 Skill:借助 pontx-hub 查询实时契约,并通过 @pontx/massive 或 pontx-massive 产品 CLI 读取 bars、snapshot、ticker 等数据,同时给出 ET/UTC 时区语义、复权口径、缺失区间、分页续传、限流重试与授权合规指引;不含凭据、账户授权与数据集本身。

pontjs
热度 11.45
最近更新 2026-08-15 17:58:09
分类:开发依赖:代码执行 / 第三方 API应用场景:API 集成 / 数据处理 / 财务 / 工作流

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add pontx-massive
Metadata
name
pontx-massive
description
Integrate Massive market data through Pontx. Use for bars, snapshots, ticker lookup, backtests, dashboards, entitlements, pagination, market-session time zones, throttling, or licensing.

Massive market data

Use this skill for product-specific market-data workflow and rights decisions. Retrieve current API and package facts from Pontx rather than copying them here.

Resolve the live contract

pontx-hub search "<market-data task>" --type endpoint --json
pontx-hub show <returned-resource-id>
pontx-hub sdk massive

Confirm the account plan, required recency, history window, and intended use before promising that data is available. Plan tiers differ in history, recency, request allowance, and dataset access; a valid key is not universal entitlement.

Use the cataloged REST surface for bounded, on-demand queries. Massive recommends Flat Files for bulk history and WebSockets for continuous live streams. Do not design a high-frequency REST polling loop for streaming work.

Preserve market-time semantics

  • Treat Massive's U.S. stock aggregate intervals as Eastern Time (ET). Interpret returned Unix timestamps as UTC using the unit in the current Schema. Implement ET with the America/New_York IANA zone rather than a fixed offset across daylight-saving transitions.
  • Make adjusted versus unadjusted history an explicit decision. Never combine series with different adjustment policies silently.
  • An aggregate interval with no qualifying trade can be absent. Do not invent a zero-price or zero-volume bar unless the user defines a filling policy.
  • Report the as-of time, recency tier, time zone, and known gaps when results drive analysis or a user-facing dashboard.

Paginate and throttle deliberately

Inspect the current pagination contract with pontx-hub show. Choose an explicit order for reproducible ingestion. Follow the returned continuation until it is absent, but accept it only on the approved Massive API host. Keep credentials out of continuation URLs and logs.

Checkpoint and deduplicate a partially retried import. Bound concurrency to the current plan. On throttling, honor Retry-After when present or use exponential backoff with jitter; distinguish authentication, entitlement, throttling, and provider failures.

Preview, then integrate

pontx-hub massive preview <api-name> <named-options>

Hub proxy execution is disabled for this product. After preview and explicit approval, execute the unchanged read from the caller process through @pontx/massive or the pontx-massive product CLI. Keep untagged SDK methods flat; never invent a common or default controller.

Read the credential environment variable from show or sdk. Never print, persist, or place its value in source, command arguments, URLs, examples, or traces. Begin product-CLI work in dry-run mode.

Respect market-data rights

Personal or non-professional access does not grant commercial redistribution, public-display, non-display, or derived-work rights. Before caching, exporting, selling, or publicly displaying data, verify subscriber classification and written rights for the intended use. If unclear, stop before retrieval and use Massive's current terms or business licensing channel.

Market data can be delayed, incomplete, corrected, or unavailable. Do not call it guaranteed and do not turn an API result into personalized investment advice.

Few-shot workflows

Scenario 1: Reproducible backtest

User: "Build an adjusted five-minute backtest across a U.S. daylight-saving transition."

Approach: Apply the market-time, adjustment, missing-interval, and pagination workflow above after inspecting the live contract and plan coverage. Preview before retrieval and preserve the audit context.

Scenario 2: Live dashboard

User: "Refresh hundreds of stock prices every second on a free plan."

Approach: Apply the entitlement, transport, and display-rights decisions above. Inspect the live plan before proposing a bounded bootstrap or a separate streaming integration.

Scenario 3: Paid public export

User: "Cache all data and expose it from our paid public API."

Approach: Stop before retrieval. Ask for business, redistribution, display, storage, and derived-work rights. Never request the key value or treat technical read access as permission to redistribute.