# AgentTV ARV (remote · mcp.agenttv.com)

Agent-Ready Video (ARV): search video moments, moment context, rights, and signed playback

- Trust score: 70/100 (medium)
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-11-05

## Components

- remote · `mcp.agenttv.com`: 70/100 (this document), [markdown](https://verifymcp.io/servers/com-agenttv-arv/c-publisher-mcp.md), [page](https://verifymcp.io/servers/com-agenttv-arv/c-publisher-mcp)

## Channel facts

- Endpoint: `https://mcp.agenttv.com/c/{publisher}/mcp`
- Transports: `streamable-http`
- Auth: `none`
- Version: `1.0.0`

## 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 2026-11-05.

- **Endpoint Security**: 74/100
  - 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 is enforced; there's no plaintext access path.
  - HSTS check failed: the Strict-Transport-Security header is absent.
  - 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**: 74/100
  - 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).
  - AI-judged instruction clarity (good).
  - Context-footprint check failed: tool/resource definitions use about 2196 tokens (~168/item across 13 items; 9 tools + 4 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 3/100
  - Stability check failed: schema churn in the 3 days we've observed: 0 tool removals, 1 breaking changes, 0 auth/transport breaks, 0 additions.
- **Tool Coverage**: 89/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 63% 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 9 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 11 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 100/100
  - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.
  - Supports UI / widget rendering.

## Install

### How do I install the AgentTV ARV MCP server?

AgentTV ARV is a hosted endpoint at https://mcp.agenttv.com/c/%7Bpublisher%7D/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 com-agenttv-arv 'https://mcp.agenttv.com/c/%7Bpublisher%7D/mcp'
```

### Cursor

```json
{
  "mcpServers": {
    "com-agenttv-arv": {
      "url": "https://mcp.agenttv.com/c/%7Bpublisher%7D/mcp"
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "com-agenttv-arv": {
      "type": "http",
      "url": "https://mcp.agenttv.com/c/%7Bpublisher%7D/mcp"
    }
  }
}
```

### Codex

```toml
[mcp_servers.com-agenttv-arv]
url = "https://mcp.agenttv.com/c/%7Bpublisher%7D/mcp"
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "com-agenttv-arv": {
      "type": "remote",
      "url": "https://mcp.agenttv.com/c/%7Bpublisher%7D/mcp",
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add com-agenttv-arv --url 'https://mcp.agenttv.com/c/%7Bpublisher%7D/mcp' --transport streamable-http
```

### Hermes

```yaml
mcp_servers:
  com-agenttv-arv:
    url: "https://mcp.agenttv.com/c/%7Bpublisher%7D/mcp"
```

### Netclaw

```json
{
  "McpServers": {
    "com-agenttv-arv": {
      "Transport": "http",
      "Url": "https://mcp.agenttv.com/c/%7Bpublisher%7D/mcp"
    }
  }
}
```

### Vellum

```bash
assistant mcp add com-agenttv-arv -t streamable-http -u 'https://mcp.agenttv.com/c/%7Bpublisher%7D/mcp'
```

### Other

```json
{
  "mcpServers": {
    "com-agenttv-arv": {
      "type": "http",
      "url": "https://mcp.agenttv.com/c/%7Bpublisher%7D/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.

### 2026-11-05 (score 70, 0)

- [security] Resource “AgentTV ARV View” now points somewhere else: ui://agenttv/arv-view.v2026110304.html → ui://agenttv/arv-view-click-to-play.v2026110304.html
- [security] The server rewrote its instructions, which are the text every model session reads
- [security] Tool “search_videos” rewrote its description, which is the text the model reads

### 2026-11-04 (score 70, 0)

- [security] Resource “AgentTV ARV View” now points somewhere else: ui://agenttv/arv-view.v2026110301.html → ui://agenttv/arv-view-click-to-play.v2026110304.html
- [security] Resource “AgentTV ARV View” now points somewhere else: ui://agenttv/arv-view-click-to-play.v2026110301.html → ui://agenttv/arv-view-click-to-play.v2026110304.html
- [security] The server rewrote its instructions, which are the text every model session reads
- [security] Tool “search_videos” rewrote its description, which is the text the model reads
- [functional regression] Schema quality: 149 → 169

### 2026-11-03 (score 70, −1)

- [security regression] Stability: unverified → fail
- [security regression] A breaking change shipped without a version bump: still 0.1.1
- [security] Resource “AgentTV ARV View” now points somewhere else: ui://agentcdn/arv-view.v2026102801.html → ui://agenttv/arv-view-click-to-play.v2026110301.html
- [security] Resource “AgentTV ARV View” now points somewhere else: ui://agentcdn/arv-view-click-to-play.v2026102801.html → ui://agenttv/arv-view-click-to-play.v2026110301.html
- [security] Resource “AgentTV ARV View” now points somewhere else: ui://agentcdn/arv-view.v2026102201.html → ui://agenttv/arv-view-click-to-play.v2026110301.html
- [security] Resource “AgentTV ARV View” now points somewhere else: ui://agentcdn/arv-view-click-to-play.v2026102201.html → ui://agenttv/arv-view-click-to-play.v2026110301.html
- [security] Resource “AgentTV ARV View” now points somewhere else: ui://agentcdn/arv-view.v2026093001.html → ui://agenttv/arv-view-click-to-play.v2026110301.html
- [security] Resource “AgentTV ARV View” now points somewhere else: ui://agentcdn/arv-view-click-to-play.v2026093001.html → ui://agenttv/arv-view-click-to-play.v2026110301.html
- [security] Resource “AgentTV ARV well-known manifest” now points somewhere else: agentcdn://protocol/arv → agenttv://protocol/arv
- [security] Resource “Integration architecture” now points somewhere else: agentcdn://protocol/integration-architecture → agenttv://protocol/integration-architecture
- [security] The server changed its declared name: agentcdn-arv → agenttv-arv
- [security] Tool “get_creator_profile” rewrote its description, which is the text the model reads
- [functional regression] Schema quality: 121 → 149
- [functional regression] “get_creator_profile” dropped the required parameter “creator_id”

### 2026-11-02 (score 71)

First indexed and scored.

## MCP tools (9)

### `search_videos` (~210 tokens)

Search videos

Finds the moments in these videos that answer a question, best first. They cover topics Autonomous AI Agents, Ember Smart Mugs, AI Video Generation; people Sander Saar, Steve Jobs, Jeff Bezos; places Los Angeles, United States. Each result has its matched text, time range, playback values (video.id, playback.segment_id, playback.start_time) and its video's languages. It returns no player and no stream URL. Every result is on topic, low band ones included. An empty result list means no match.

Input parameters:

- `cursor` (string): Optional. The next_cursor of an earlier list response, to read the next page of the same list.
- `filters` (object)
- `query` (string, required)
- `results` (string): Optional. "all" returns every matching moment grouped by video, with counts and next_cursor. "best" returns the ranked moments that best answer the query. Omitted, a query that asks for every instanc…

Output parameters:

- `featured_stream` (object)
- `list` (object)
- `results` (array)

### `search` (~85 tokens)

Search video moments

Finds the moments in this video catalog that answer a question, best first, with public citation links. Each result quotes its matched text, names its channel and carries a confidence band. Every result is on topic, low band ones included. An empty result list means no match. A result with empty text carries a text_omitted_reason that says why.

Input parameters:

- `query` (string, required)

Output parameters:

- `results` (array)

### `fetch` (~54 tokens)

Fetch video moment

Fetches the full grounded context for one search result and returns its public citation link, including on-screen text with its timecode. When it finds no evidence, text_omitted_reason says why.

Input parameters:

- `id` (string, required)

Output parameters:

- `accepts` (array)
- `error` (string)
- `extensions` (object)
- `id` (string)
- `metadata` (object)
- `refusal`
- `resource` (object)
- `text` (string)
- `text_omitted_reason` (string)
- `text_sources` (array)
- `title` (string)
- `url` (string)
- `x402Version` (number)

### `get_moment_context` (~156 tokens)

Get moment context

Returns cached transcript chapters, timed transcript sentences, shot captions, OCR, and entities for a video window without model inference. Input times use seconds from the start of the video. Response fields ending in _ms use milliseconds. Omitted input times return the first window plus the next start time. An empty section carries its reason under omitted. granularity "word" adds the start and end time of each word, as for a clip cut.

Input parameters:

- `end_time` (number)
- `granularity` (string): Optional. "sentence" (default) returns timed sentences. "word" also returns word-level start and end times for the window.
- `start_time` (number)
- `video_id` (string, required)

Output parameters:

- `chapters` (array)
- `entities` (array)
- `labels` (array)
- `model_provenance` (array)
- `omitted` (object)
- `pagination` (object)
- `parts` (array)
- `people` (array)
- `points` (array)
- `products` (array)
- `shots` (array)
- `transcript_sentences` (array)
- `truncated` (object)
- `verbosity_applied` (string)
- `video` (object)
- `word_timings` (object)

### `get_creator_profile` (~36 tokens)

Get creator profile

Returns this server's creator identity, authority signals, namespace, and catalog summary. It takes no arguments and does not return playback data.

Output parameters:

- `allow_search_indexing` (boolean)
- `authority_signals` (object)
- `bio` (string)
- `catalog_summary` (object)
- `discoverable` (boolean)
- `id` (string)
- `name` (string)
- `namespace_url` (string)
- `network_carriage` (string)
- `platforms` (array)
- `profile_image_url` (string)
- `slug` (string)
- `verified` (boolean)
- `verified_via` (string)

### `get_video_stream` (~236 tokens)

Get signed video stream

Returns a signed playback response for one selected video moment and supplies the inline playback card. A segment_id, start_time, or query provides the required moment anchor. A query can rerank the moment and populate alternatives. A video with no moment for the anchor returns no match. The video names its spoken and subtitle languages (language_primary, caption_languages) when known.

Input parameters:

- `query` (string): The user's question. Re-ranks the moment within this video and fills the alternatives row.
- `segment_id` (string): Required when neither start_time nor query is given. The playback.segment_id of the chosen search result, exactly as given. It names the moment to play.
- `start_time` (number): Required when neither segment_id nor query is given. Seconds into the video, as in the chosen search result's playback.start_time. 0 is the top of the video. When a query is also present, a 0 is trea…
- `video_id` (string, required): The video's id, exactly as a search result gives it (video.id, or asset_id).

Output parameters:

- `accepts` (array)
- `alternatives` (array)
- `arv` (string)
- `asset_id` (string)
- `creator` (object)
- `delivery_kind` (string)
- `embed_url` (string)
- `end_time` (number)
- `error` (string)
- `expires_at` (string)
- `extensions` (object)
- `id` (string)
- `match` (object)
- `media_url` (string)
- `moment_uri` (string)
- `presentation` (object)
- `preview` (object)
- `pst` (string)
- `refusal`
- `resource` (object)
- `share_url` (string)
- `source_name` (string)
- `source_page_url` (string)
- `start_time` (number)
- `stream_url` (string)
- `watch_full_url` (string)
- `watch_full_url_at_moment` (boolean)
- `x402Version` (number)

### `get_segment_rights` (~70 tokens)

Get segment rights

Returns evaluated usage rights for quoting, clipping, AI training, licensing, and playback. Search and playback tools enforce applicable rights independently. The default destination modality is generic_mcp.

Input parameters:

- `destination_modality` (string)
- `segment_id` (string)
- `video_id` (string, required)

Output parameters:

- `agent_identity_scope` (array)
- `ai_training_allowed` (boolean)
- `commercial_use` (boolean)
- `evaluated_at` (string)
- `evaluated_for` (object)
- `expires_at` (string)
- `permissions` (object)
- `policy_id` (string)
- `policy_version` (integer)
- `requires_attribution` (boolean)
- `requires_entitlement`
- `revocable` (boolean)
- `subscription_tier_required` (string|null)
- `territory` (string)
- `time_window`
- `user_type` (string)

### `record_usage` (~381 tokens)

Record usage

Records one use of a creator's video moment (shown, quoted, cited, summarized or played) and returns a signed UsageReceipt. The moment is named by moment_uri, or by asset_id with start_ms and end_ms. action is a rights term such as cite, quote or play. A use the rights policy forbids is refused and logged. served_to names the agent and surface the moment was served to. The receipt names the verified agent, or unverified. A repeat with the same idempotency_key returns the original receipt.

Input parameters:

- `action` (string, required): What the agent did with the moment: one term of the RightsSummary, for example cite, quote or segment_display. A term the rights do not define is refused.
- `asset_id` (string): A video id. With start_ms and end_ms it names any range on a published video.
- `end_ms` (number): Range end in whole milliseconds.
- `idempotency_key` (string): Up to 128 characters A-Z, a-z, 0-9, dot, underscore, colon or hyphen, for example a UUID. One key identifies one use. A repeat with the same key within the same UTC hour returns the original receipt…
- `moment_id` (string): The moment id of a search result (mom_ followed by 26 characters a-z and 2-7), checked against the range. It cannot name a moment alone.
- `moment_uri` (string): The moment's ARV URI, arv:{asset_id}#t={start_ms},{end_ms}, as a search result gives it.
- `served_to` (object): Who the moment was served to: the agent name and the surface, for example chatgpt and chat.
- `start_ms` (number): Range start in whole milliseconds.

Output parameters:

- `action` (string)
- `arv` (string)
- `citation` (string)
- `entitlement_ref` (string)
- `jws` (string)
- `moment_id` (string)
- `moment_uri` (string)
- `payment_ref` (string)
- `policy_version` (integer)
- `receipt_id` (string)
- `served_at` (string)
- `served_to` (object)

### `get_memory_page` (~163 tokens)

Get memory page

Returns a cited summary page of this creator's library: kind library lists the topics and people it covers, and kind video, topic or entity (with subject) reads one. Every line cites video ranges by video_id, start_time and end_time. Text and citations only, no playback. found is false when the library has no such page yet.

Input parameters:

- `kind` (string): Which page to read. library is this creator's index of topic and entity pages (the default). video, topic and entity read one subject.
- `subject` (string): The page's subject. For video: a video id. For topic: a topic_id, and for entity: an entity_id, both as the refs of a page line give them. Not used for library.

Output parameters:

- `compiler_version` (string)
- `found` (boolean)
- `kind` (string)
- `reason` (string)
- `sections` (array)
- `subject` (string)
- `title` (string)

## Diagnostics

Captured diagnostic sections: TLS, DNSSEC, Authorisation, Transports. The full working is on the page: https://verifymcp.io/servers/com-agenttv-arv/c-publisher-mcp#diagnostics

## Score history

- 2026-11-05: 70
- 2026-11-04: 70
- 2026-11-03: 70
- 2026-11-02: 71

## Common questions

### What is the AgentTV ARV MCP server?

AgentTV ARV is an MCP server listed in the public MCP registry as com.agenttv/arv. Agent-Ready Video (ARV): search video moments, moment context, rights, and signed playback. This page covers its hosted endpoint (https://mcp.agenttv.com/c/{publisher}/mcp).

### Is the AgentTV ARV MCP server safe to use?

AgentTV ARV scores 70 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 AgentTV ARV MCP server expose?

AgentTV ARV exposes 9 tools: search_videos, search, fetch, get_moment_context, get_creator_profile, and 4 more. Their descriptions and schemas cost roughly 1,391 tokens of context every time the server is loaded.

### Does the AgentTV ARV MCP server require authentication?

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

### Is the AgentTV ARV MCP server still maintained?

AgentTV ARV is still listed as active in the MCP registry. We last reached this channel on 5 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 AgentTV ARV MCP server?

VerifyMCP ran 14 checks on AgentTV ARV, covering TLS certificate, Authorization, HTTPS, HSTS header and 10 more. 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://mcp.agenttv.com/c/%7Bpublisher%7D/mcp
- Website: https://agenttv.com/
- Changelog RSS feed: https://verifymcp.io/servers/com-agenttv-arv/c-publisher-mcp.xml
- Changelog JSON feed: https://verifymcp.io/servers/com-agenttv-arv/c-publisher-mcp/changelog.json
- Trust report (JSON): https://verifymcp.io/servers/com-agenttv-arv/c-publisher-mcp.json
- HTML version of this page: https://verifymcp.io/servers/com-agenttv-arv/c-publisher-mcp