canonical: https://jentic.com/apis/newruntime.com/newruntime

# New Runtime public JSON and MCP API

The New Runtime public JSON API publishes New Runtime's AI-industry research as typed, read-only JSON: raw signals, patterns, topic hubs, field notes, capability deltas, structural shifts, system dossiers with release timelines, project blueprints, and Learn guides, each listable and readable by slug. It also exposes the site's own machine-readership data, such as bot traffic aggregates, crawler journeys, and a discovery graph, and a free operation that contextualizes an outside URL against published records. A moderated agent inbox accepts sources, corrections, and usage receipts, and optional x402 paid tools find the primary source behind a claim, check an AI-industry rumour against X discussion, test for prior coverage, and build research bundles.

## For AI agents

Read New Runtime's AI-industry signals, patterns, shifts, and system dossiers as JSON, contextualize a URL against them, submit moderated corrections, and call x402 paid tools for primary sources and rumour checks.

## Scope

Does not provide general web search, live market data, or posting to social networks. Use for reading New Runtime's AI-industry research and its verification tools only.

## Capabilities

- List and read AI-industry signals, patterns, topic hubs, field notes, capability deltas, and structural shifts by slug
- Read a system dossier and its release timeline, or a project blueprint, as structured JSON
- Contextualize an outside https URL against published records, related patterns, and its source hub for free
- Submit a source, a correction with a problem of outdated, unsupported, contradicted, or other, or a usage receipt to the moderated agent inbox and check its moderation status
- Find the official primary source behind a post, article, or claim through an x402 paid tool that settles only when a source is found
- Check an AI-industry rumour against recent X discussion and official statements, returning verdicts such as corroborated or official_source_found
- Read the site's public bot traffic, crawler journey, page readership, and discovery graph data

## Use cases

### Research agent grounding AI-industry answers

A research agent asked about a new AI tool or model wants structured, sourced records rather than a general web search. Through Jentic the agent searches for 'contextualize a URL against AI-industry research', loads the New Runtime contextualize operation with its input schema, and passes the url and an optional task. It gets back records citing that URL, related records and patterns, and its source hub, then reads individual signals or system dossiers by slug for detail.

Example prompt: Call GET `/agent-actions/contextualize.json` with url 'https://github.com/openai/codex' and task 'summarize coverage', then read the first related record with GET `/signals/{slug}.json`

### Verifying a claim before repeating it

Before an agent repeats an AI-industry rumour, it checks it. The rumour-check tool searches X over the last 7 days with reposts and replies excluded, counts independent authors and trusted voices, and looks for an official statement, returning a verdict such as official_source_found, corroborated, discussed_unconfirmed, or no_discussion. The unsigned request validates input for free and the check runs after x402 payment verification.

Example prompt: Send POST `/agent-payments/rumor-check` with claim 'Vendor X is deprecating its v1 API next month', review the 402 quote, and retry with a PAYMENT-SIGNATURE only if the user approves

### Feeding corrections back to the publisher

When an agent finds a published claim that a newer source contradicts, it can report it instead of silently ignoring it. The correction operation takes the record, the claim, a problem such as outdated or contradicted, an evidence_url, and the submitting agent, and queues it for weekly moderated review without changing the site. The submission's moderation status can be checked later by id.

Example prompt: Call POST `/agent-inbox/correction` with record, claim, problem 'outdated', evidence_url, and agent, then poll GET `/agent-inbox/submissions/{submission_id}.json`

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| GET | `/signals.json` | List raw signals |
| GET | `/systems/{slug}.json` | Read a system dossier |
| GET | `/systems/{slug}/releases.json` | Read a system release timeline |
| GET | `/agent-actions/contextualize.json` | Contextualize an outside URL |
| POST | `/agent-payments/rumor-check` | Check an AI-industry rumour |
| POST | `/agent-payments/primary-source` | Find the primary source behind a claim |
| POST | `/agent-inbox/correction` | Submit a correction |
| GET | `/feed.json` | Read the JSON Feed |

## Key resources

- **Research records** — Signals, patterns, topics, field notes, deltas, shifts, graveyard records, projects, systems, and Learn guides, listed and read by slug
- **Agent actions** — Free URL contextualization against published records
- **Agent inbox** — Moderated submissions of sources, corrections, and usage receipts with status lookup
- **Paid tools** — x402 primary-source finder, rumour check, prior-coverage check, and research bundles with quote previews
- **Machine readership** — Bot traffic, crawler journeys, page readership, discovery graph, and source ledger data

## AI readiness

This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use.

- **Score:** 70 / 100
- **Maturity:** AI-Aware
- **Dimensions:**
  - Foundational Compliance: 51 / 100
  - Developer Experience & Jentic Compatibility: 68 / 100
  - AI-Readiness & Agent Experience: 69 / 100
  - Agent Usability: 91 / 100
  - Security: 100 / 100
  - AI Discoverability: 60 / 100
- **View full report:** https://jentic.com/apis/newruntime.com/newruntime/scorecard
- **How the score is calculated:** https://docs.jentic.com/reference/api-readiness-framework/overview/
- **More about the dimensions:** https://docs.jentic.com/reference/api-readiness-framework/specification/#dimensional-model-overview

### Score it yourself

Every API in the directory is allowlisted, so you can re-score it with no key required.

- **Score your own API:** https://jentic.com/scorecard.md
- **Scoring CLI agent skill:** https://github.com/jentic/jentic-api-scorecard/blob/main/skills/jentic-api-scorecard/SKILL.md

```sh
npx @jentic/api-scorecard-cli score <openapi-url>
```

## Why Jentic

- **Setup:** Wiring New Runtime by hand means mapping dozens of static JSON routes, telling free reads apart from x402 paid routes that answer an unsigned POST with a 402 quote, and pacing calls against the RateLimit fields. Through Jentic you install once, import New Runtime from the API Directory, and your agent calls it with no key to configure.
- **Permission scoping:** New Runtime puts the record slug in the URL path, but the bound that matters is by operation and path prefix: a rule can hold your agent to the free GET reads. You choose the operations it may call, so the paid routes under /agent-payments and inbox submissions are not included unless you add them.
- **Credential handling:** New Runtime declares no API credential, so there is no key for your Jentic One instance to hold for the free routes. Paid routes are settled per call with an x402 PAYMENT-SIGNATURE rather than an account key.
- **Discovery method:** Agents search Jentic by intent such as 'check an AI rumour' or 'read a system dossier', and Jentic returns the matching New Runtime operation with its input schema so the agent picks the right route among 68.

## Related APIs

- **News API** — Broad news article search across publishers
- **Hacker News API** — Live developer discussion and story feeds
- **Feedly API** — Feed reading and aggregation across many sources

## FAQ

### What authentication does the New Runtime API use?

The OpenAPI spec declares no security schemes, and every operation is public. Free reads need no account or API key. The paid tools also need no account: the first unsigned POST validates the request and, for a sellable result, returns an HTTP 402 x402 quote, and you retry with a PAYMENT-SIGNATURE header to settle it.

### Can I read the release history of an AI system with the New Runtime API?

Yes. GET /systems.json lists system dossiers, GET `/systems/{slug}.json` returns one structured dossier with private commands and credentials excluded, and GET `/systems/{slug}/releases.json` returns its release timeline.

### What are the rate limits for the New Runtime API?

The spec states that public static JSON reads enforce 120 requests per minute and return live RateLimit fields in the response, so an agent can pace itself from those headers.

### Is there a New Runtime MCP server?

Yes - New Runtime hosts a read-only MCP endpoint at POST /mcp. You can also connect the API directly through Jentic: your agent gets the full 68-endpoint surface from the spec without loading another server's tool definitions into its context, and your rules decide which operations the agent may call, including whether it can reach the paid routes.

### How do I contextualize a URL with New Runtime through Jentic?

Import New Runtime from the API Directory in your Jentic One instance; there is no key to add. Your agent searches for 'contextualize a URL', loads GET `/agent-actions/contextualize.json`, and passes the https url plus an optional task. To run it on your own infrastructure, install Jentic One from its GitHub repo.

### Can I limit what my agent is allowed to do with the New Runtime API?

Yes. Write a rule that allows only the GET operations, and the agent can read every research record and contextualize URLs but cannot call the x402 paid tools under /agent-payments or post to the agent inbox. Every call it makes is logged.