# Timefire (remote · api.timefire.ai)

Timefire's public record of AI stock forecasts: graded books, track record and proof.

- Trust score: 75/100 (medium)
- Change this week: +3
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-11-21

## Components

- remote · `api.timefire.ai`: 75/100 (this document), [markdown](https://verifymcp.io/servers/ai-timefire-record/api.md), [page](https://verifymcp.io/servers/ai-timefire-record/api)

## Channel facts

- Endpoint: `https://api.timefire.ai/mcp/`
- Transports: `streamable-http`
- Auth: `none`
- Version: `0.1.1`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically against the live server, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-11-21.

- **Endpoint Security**: 69/100
  - The endpoint's TLS certificate is valid, in date, and uses a strong key.
  - No authorisation is required to call this server. Nothing on it is marked destructive, but we can't show that an anonymous caller can only read: full credit needs the server to list at least one tool and every tool to be verifiably read-only. We use a best-effort check of tool names and descriptions to tell read-only tools from tools that change things, so some genuinely read-only tools may not get full credit. Names that start with a read verb (get, list, read, search and similar) and accurate readOnlyHint annotations help. Meet that to earn full credit, or require authorisation.
  - HTTPS enforcement could not be verified: the plaintext port answered with HTTP 406, which proves neither a plaintext path nor enforcement.
  - The HSTS (Strict-Transport-Security) header is present.
  - DNSSEC check failed: this domain isn't protected by DNSSEC.
- **Transport & Reachability**: 100/100
  - Verified streamable-http transport via a live MCP handshake.
- **Schema Quality & AI Usability**: 69/100
  - AI-judged instruction clarity (excellent).
  - Context-footprint check failed: tool/resource definitions use about 1937 tokens (~161/item across 12 items; 12 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 47/100
  - Stability observed for 14 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 100/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 100% of tool parameters carry a description.
  - Structured output schemas are declared (100% of tools); any adoption earns full credit.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - We read all 12 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 13 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 100/100
  - Implements a current MCP spec version (2026-07-28).

## Install

### How do I install the Timefire MCP server?

Timefire is a hosted endpoint at https://api.timefire.ai/mcp/, so there is nothing to install locally. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

### Claude

```bash
claude mcp add --transport http ai-timefire-record 'https://api.timefire.ai/mcp/'
```

### Cursor

```json
{
  "mcpServers": {
    "ai-timefire-record": {
      "url": "https://api.timefire.ai/mcp/"
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "ai-timefire-record": {
      "type": "http",
      "url": "https://api.timefire.ai/mcp/"
    }
  }
}
```

### Codex

```toml
[mcp_servers.ai-timefire-record]
url = "https://api.timefire.ai/mcp/"
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "ai-timefire-record": {
      "type": "remote",
      "url": "https://api.timefire.ai/mcp/",
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add ai-timefire-record --url 'https://api.timefire.ai/mcp/' --transport streamable-http
```

### Hermes

```yaml
mcp_servers:
  ai-timefire-record:
    url: "https://api.timefire.ai/mcp/"
```

### Netclaw

```json
{
  "McpServers": {
    "ai-timefire-record": {
      "Transport": "http",
      "Url": "https://api.timefire.ai/mcp/"
    }
  }
}
```

### Vellum

```bash
assistant mcp add ai-timefire-record -t streamable-http -u 'https://api.timefire.ai/mcp/'
```

### Other

```json
{
  "mcpServers": {
    "ai-timefire-record": {
      "type": "http",
      "url": "https://api.timefire.ai/mcp/"
    }
  }
}
```

The mcpServers block is a cross-client convention. Remote transports vary, so check your client's docs.

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-11-21 (score 75, +1)

No change was recorded against any check on this day. Stability & Change Management went from 43 to 47. That category is still filling its 30-day observation window: 13 days of observed history at the previous scan, 14 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-19 (score 74, +1)

No change was recorded against any check on this day. Stability & Change Management went from 37 to 40. That category is still filling its 30-day observation window: 11 days of observed history at the previous scan, 12 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-17 (score 73, +1)

No change was recorded against any check on this day. Stability & Change Management went from 30 to 33. That category is still filling its 30-day observation window: 9 days of observed history at the previous scan, 10 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-15 (score 72, 0)

- [security] Tool “verify_batch” rewrote its description, which is the text the model reads
- [functional regression] Schema quality: 145 → 161

### 2026-11-14 (score 72, +1)

No change was recorded against any check on this day. Stability & Change Management went from 20 to 23. That category is still filling its 30-day observation window: 6 days of observed history at the previous scan, 7 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-12 (score 71, +3)

- [security improvement] HSTS header: fail → pass

### 2026-11-11 (score 68, +1)

No change was recorded against any check on this day. Stability & Change Management went from 10 to 13. That category is still filling its 30-day observation window: 3 days of observed history at the previous scan, 4 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-09 (score 67, +1)

No change was recorded against any check on this day. Stability & Change Management went from 3 to 7. That category is still filling its 30-day observation window: 1 days of observed history at the previous scan, 2 at this one. The score rises as the window fills, whether or not the server changes.

## MCP tools (12)

### `list_resolved_batches` (~159 tokens)

Book history

List graded books, newest first, one summary row each: date, returns, alpha against SPY and RSP, and verdict.
    A book (or batch) is one trading day's full S&P 500 slate. Use this to find a trading_date, then get_resolved_batch
    for every forecast in that book or verify_batch for its proof. Dates with no graded book return an empty list.

Input parameters:

- `from_date`: Earliest date to include, as YYYY-MM-DD. Omit for no lower bound.
- `limit` (integer): Most books to return, 1 to 500. Default 60.
- `to_date`: Latest date to include, as YYYY-MM-DD. Omit for no upper bound.

Output parameters:

- `result` (array)

### `get_resolved_batch` (~124 tokens)

One book

Get one graded book with every forecast and its outcome. A book (or batch) is one trading day's full S&P 500
    slate, served once its long picks are fully graded, so other rows can still have null outcomes. Use
    list_resolved_batches to find the date, verify_batch for the book's proof and get_book_portfolio for its portfolio.
    An ungraded or unknown date is an error.

Input parameters:

- `trading_date` (string, required): The book's trading date, as YYYY-MM-DD. list_resolved_batches gives the graded dates.

### `get_track_record_stats` (~136 tokens)

Headline stats

Get aggregates computed only from fully resolved books: forecast counts, returns and hit rates, and how many
    books beat, matched (flat) or trailed SPY, each book's verdict graded on its cap-weighted long picks. The hit
    rates count the books graded under the published rule, from published_rule_from on; the 69 backtest books before
    it follow a rule of their own and are reported apart in the backtest_ fields, verdicts included. Use this for
    overall performance, get_calibration to test the ranking itself, and get_book_portfolio for a single book.

### `get_calibration` (~80 tokens)

Calibration

Check whether forecasts matched outcomes: mean predicted against mean realized return by decile across every
    graded book, and hit rates by rank bucket and by side (long or short) across books graded under the published rule.
    Use this to judge how well the ranking held up, and get_track_record_stats for headline returns and the verdict
    tally.

### `get_universe_and_scoring_spec` (~56 tokens)

Record spec

Get the versioned scoring rules, the metric definitions and the latest graded book's universe of stocks.
    Use this before interpreting a stat or a verdict, to learn how books are graded and which stocks are covered.

### `get_stock_resolved_history` (~121 tokens)

One stock's history

Get one stock's rows from graded books only, newest first: its forecast, rank, pick side and outcome in each
    book. Use this to follow one ticker across books, explain_resolved_prediction for the evidence behind one row, and
    get_resolved_batch for a whole book. An unknown ticker returns an empty list.

Input parameters:

- `limit` (integer): Most rows to return, 1 to 500. Default 120.
- `ticker` (string, required): Stock ticker, for example AAPL. Upper or lower case both work.

Output parameters:

- `result` (array)

### `verify_batch` (~316 tokens)

Verification record

Get the public proof for one graded book: signed payload, hash, signature, content address and timestamp
    material, with the steps to check them. From 2026-04-28 on, a book is made after the close on its trading date
    (earlier books are a backtest, built later), and its return runs from that close over the next ten NYSE sessions
    (books before 2026-10-21 run from the first available close on or after that date over the next ten available
    sessions; the backtest books dated 2026-01-02 to 2026-04-13 run from the close before their date to nine sessions
    after it). The OpenTimestamps receipt proves the book existed before the Bitcoin block it names: compare that
    block's time with the next NYSE open after the trading date. sealed is true only while a book is still withheld.
    Every book served here is graded and revealed, so sealed is false, and that is not a failed check. The verified
    field is the server's own check of the hash and signature, not the timestamp; the evidence lets you repeat it
    yourself. Use this to check a book's signature and timestamp, and get_resolved_batch to read its forecasts and
    outcomes. An ungraded or unknown date is an error.

Input parameters:

- `trading_date` (string, required): The book's trading date, as YYYY-MM-DD. list_resolved_batches gives the graded dates.

### `get_book_portfolio` (~216 tokens)

Book portfolio

Get a resolved book's portfolio: market-cap-weighted vs SPY ("cap", the default, the portfolio the ledger's
    verdict grades) or equal-weighted vs RSP ("equal"), with each long pick's weight and graded return, the book
    return, the benchmark return and its evidence, alpha and verdict. Equal weights always pair with RSP and cap weights
    with SPY. Per prediction, books serve the two as equal_weight and market_cap_weight; never read a pick's cap_weight
    as either without checking inference_metadata.weighting. Use this for one book's result, and get_simulation_nav for
    the running portfolio across books. An ungraded or unknown date is an error.

Input parameters:

- `trading_date` (string, required): The book's trading date, as YYYY-MM-DD. list_resolved_batches gives the graded dates.
- `weighting` (string): "cap" for market-cap weights against SPY (the default), or "equal" for equal weights against RSP.

### `search_docs` (~116 tokens)

Search the docs

Find the developer guides and site pages that match a few words. Each
    hit carries the page title, its canonical URL and a short excerpt. Use this
    for how Timefire works and what its terms mean, not for record data. No
    match returns an empty list, so try other words.

Input parameters:

- `limit` (integer): Most pages to return, 1 to 20. Default 5.
- `query` (string, required): A few words to search for, at most 8 words and 200 characters.

Output parameters:

- `result` (array)

### `get_simulation_nav` (~185 tokens)

Portfolio value

Get the paper portfolio's daily value (NAV), market-cap-weighted vs SPY ("cap", the default) or equal-weighted
    vs RSP ("equal"), with both benchmarks on every row. There is one row per resolved trading day. The series stops at
    the newest date by which every book is revealed, because a later value holds positions from books still sealed.
    Use this for portfolio value over time, and get_simulation_trades for the trades behind it.

Input parameters:

- `from_date`: Earliest date to include, as YYYY-MM-DD. Omit for no lower bound.
- `to_date`: Latest date to include, as YYYY-MM-DD. Omit for no upper bound.
- `weighting` (string): "cap" for market-cap weights against SPY (the default), or "equal" for equal weights against RSP.

Output parameters:

- `result` (array)

### `get_simulation_trades` (~214 tokens)

Portfolio trades

Get the paper portfolio's closed trades from resolved books, newest first, market-cap-weighted ("cap", the
    default) or equal-weighted ("equal"): the two trade at the same prices and differ in weight, and so in size and
    P&L. Open positions never appear. Date bounds apply to the day a trade opened; narrow them to see past the limit.
    Use this for individual trades, and get_simulation_nav for portfolio value.

Input parameters:

- `from_date`: Earliest date to include, as YYYY-MM-DD. Omit for no lower bound.
- `limit` (integer): Most trades to return, 1 to 500. Default 500. Older trades are cut off; there is no paging.
- `to_date`: Latest date to include, as YYYY-MM-DD. Omit for no upper bound.
- `weighting` (string): "cap" for market-cap weights against SPY (the default), or "equal" for equal weights against RSP.

Output parameters:

- `result` (array)

### `explain_resolved_prediction` (~144 tokens)

Explain a prediction

Get the stored evidence behind one stock's forecast in a graded book: its rank, side, market regime, 5- and
    20-day lookback returns, news items and notes. Use this to see the context a stock was ranked in, and
    get_stock_resolved_history to find which books hold it. An ungraded date, or a ticker not in that book, is an
    error.

Input parameters:

- `ticker` (string, required): Stock ticker, for example AAPL. Upper or lower case both work.
- `trading_date` (string, required): The book's trading date, as YYYY-MM-DD. list_resolved_batches gives the graded dates.

## Diagnostics

Captured diagnostic sections: TLS, DNSSEC, Authorisation, Transports. The full working is on the page: https://verifymcp.io/servers/ai-timefire-record/api#diagnostics

## Score history

- 2026-11-21: 75
- 2026-11-20: 74
- 2026-11-19: 74
- 2026-11-18: 73
- 2026-11-17: 73
- 2026-11-16: 72
- 2026-11-15: 72
- 2026-11-14: 72
- 2026-11-13: 71
- 2026-11-12: 71
- 2026-11-11: 68
- 2026-11-10: 67
- 2026-11-09: 67
- 2026-11-08: 66
- 2026-11-07: 62

## Common questions

### What is the Timefire MCP server?

Timefire is an MCP server listed in the public MCP registry as ai.timefire/record. Timefire's public record of AI stock forecasts: graded books, track record and proof. This page covers its hosted endpoint (https://api.timefire.ai/mcp/).

### Is the Timefire MCP server safe to use?

Timefire scores 75 out of 100 on VerifyMCP. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

### What tools does the Timefire MCP server expose?

Timefire exposes 12 tools: list_resolved_batches, get_resolved_batch, get_track_record_stats, get_calibration, get_universe_and_scoring_spec, and 7 more. Their descriptions and schemas cost roughly 1,867 tokens of context every time the server is loaded.

### Does the Timefire MCP server require authentication?

No. We connected to Timefire without credentials and it answered, so anything it exposes is reachable by anyone who knows the address.

### Is the Timefire MCP server still maintained?

Timefire is still listed as active in the MCP registry. We last reached this channel on 21 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

### What did VerifyMCP test on the Timefire MCP server?

VerifyMCP ran 12 checks on Timefire, covering TLS certificate, Authorization, HSTS header, DNSSEC and 8 more. One item could not be completed, a gap on our side rather than a finding about Timefire: HTTPS. Supply Chain Security and Provenance & Transparency do not apply to a hosted endpoint, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.

## Links

- Remote endpoint: https://api.timefire.ai/mcp/
- Repository: https://github.com/timefire-ai/timefire-plugin
- Website: https://timefire.ai/developers/mcp/
- Changelog RSS feed: https://verifymcp.io/servers/ai-timefire-record/api.xml
- Changelog JSON feed: https://verifymcp.io/servers/ai-timefire-record/api/changelog.json
- Trust report (JSON): https://verifymcp.io/servers/ai-timefire-record/api.json
- HTML version of this page: https://verifymcp.io/servers/ai-timefire-record/api
