Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@com.losercoreplaylist/creative-tools Prepare a melancholy three-song video draft with captions and a workspace link."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Public integration tools for Losercore Playlist: starter-track discovery, original scene prompts and editable playlist-video drafts. The browser workspace supports free local editing with your own audio and text/LRC.
Streamable HTTP: https://losercoreplaylist.com/mcp
Authentication: none. No API keys or environment variables.
claude mcp add --transport http losercore-playlist https://losercoreplaylist.com/mcpFor other clients, add the URL as a remote Streamable HTTP MCP server. Setup and feature guide.
Official MCP Registry record: com.losercoreplaylist/creative-tools v1.0.0. server.json describes this existing remote service; it is not an npm package manifest.
Related MCP server: Sonicmatch
npx skills add nullurl/losercore-agent-tools --skill losercore-playlistClawHub listing · Download Skill ZIP
Read the skill first. It contains instructions, public reference links and a dependency-free Node.js request helper. It supports agents implementing the SKILL.md format, including Codex and Claude Code. The skill does not automatically connect an MCP server; use the connection above or its bundled helper.
Try: “Prepare a melancholy three-song Losercore playlist-video draft in Chinese. Use original captions and give me the JSON to import into the free video workspace.”
Tool | Inputs | Result |
| query, limit | Song/artist/mood metadata from four starter tracks, dated 2026-10-21 |
| beanie-loser or turn-on-the-playlist | Original adult-character visual scene prompt |
| 1–12 title/artist pairs; optional three original captions and en/zh/id language | Editable JSON draft and browser workspace URL |
Two resources expose starter metadata and workspace guidance; plan_playlist_video supplies a mood-based planning prompt. Discover current input schemas through tools/list. A live Spotify catalogue, music files and song lyrics are not provided.
Call prepare_video_project with song titles/artists and three short original captions, or use its defaults.
Save only the returned project object as the returned filename.
Open workspaceUrl and use Import AI project to load the file.
Add audio/text or LRC you have permission to publish, review timing, and export locally. Save your work before refreshing. MP4 codec support depends on the browser; Chrome/Edge is recommended.
Preparing a draft is free and costs 0 credits. It does not render an MP4 or submit a cloud AI job.
This repository includes a runnable bridge for MCP clients restricted to stdio. It forwards tools, resources and prompts to the same hosted public service; it is not an offline copy of the website backend.
git clone https://github.com/nullurl/losercore-agent-tools.git
cd losercore-agent-tools
npm ci --ignore-scripts
npm startNode.js 22.12+ is required. Configure your stdio client with command: "node" and args: ["/absolute/path/losercore-agent-tools/src/stdio.mjs"]. No npm package has been published under this name. Network access to losercoreplaylist.com is required. A Dockerfile is included for directory inspection and stdio execution.
llms.txt links to interlinked Markdown references for features, pricing, data scope, site structure and MCP. llms-full.txt combines the reference material.
Local editing is free. Optional cloud AI is separate and uses credits. Fetch the live pricing reference for current amounts and purchase availability rather than assuming all displayed plans are on sale. Neither this skill nor this MCP exposes account balances, purchases, private media or cloud-generation credentials.
Tool inputs pass through your chosen agent host and the Losercore service. The bridge/helper sends no API keys, local media or account cookies. Supplied titles/captions are processed for the request; this integration does not save them as a website project. Do not send secrets or private media. This public repository contains integration tooling only; it does not contain the private website source or deployment credentials.
Privacy · Terms · Support: joyforjoker43@gmail.com
The skill directory is MIT-0 (matching ClawHub publication terms); the bridge and other root files are MIT. These licenses do not grant rights to third-party recordings, lyrics, brands, or the separately hosted website implementation.
npm test
node skills/losercore-playlist/scripts/mcp-call.mjs tools/listThe smoke test uses the official MCP SDK to connect through the local stdio bridge, lists tools/resources/prompts, prepares an original draft and confirms invalid inputs fail. It calls only the public free endpoint.
Return one of two original adult-character scene prompts for a PS2-inspired concept. This prepares text only and never submits or charges an AI generation request.
| Name | Required | Description | Default |
|---|---|---|---|
| template | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnly=true, idempotent=true, destructive=false, so the safety profile is covered. The description usefully adds that nothing is submitted or charged, which clarifies cost side-effects, but says nothing about what the returned prompt text looks like or how the two templates differ.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two sentences, zero padding, with the core action front-loaded and the no-cost/no-submit caveat immediately after. Every sentence carries distinct information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With no output schema, the description does its job by saying the result is scene prompt text prepared without a generation call. The main remaining gap is guidance on the two template values, which an agent needs to pick one correctly.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Only one parameter, an enum of 'beanie-loser' and 'turn-on-the-playlist', matching 'one of two' in the description. Schema coverage is 0% and the description does not explain what either template actually contains, so an agent cannot choose between the enum values from the definition alone.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific verb ('Return') and a concrete resource ('one of two original adult-character scene prompts for a PS2-inspired concept'), which is clear enough for an agent to distinguish it from search_playlist. It does not, however, contrast itself with the sibling prepare_video_project, so sibling differentiation is left implicit.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The line 'prepares text only and never submits or charges an AI generation request' implies this is the pre-generation step, which hints at when to use it. But there is no explicit statement of when to choose this over prepare_video_project, nor any named alternative or exclusion.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Prepare a free local-workspace JSON draft from 1–12 song/artist title cards and three original captions. Save the project field as JSON, then import it on the linked workspace. It does not render a video, fetch audio or provide song lyrics.
| Name | Required | Description | Default |
|---|---|---|---|
| tracks | Yes | ||
| captions | No | ||
| language | No | en |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true, idempotentHint=true and destructiveHint=false, so the safety profile is covered. The description adds genuinely new context: the output is a local draft, not a rendered video, and no audio or lyrics are fetched, which prevents the agent from over-promising a final artifact.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two sentences, front-loaded with the core action and artifact, followed by the operational next step and the negative boundaries. No filler.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With no output schema, the description usefully names the 'project' field that is returned and how to use it, which is the key missing piece. The unexplained 'language' parameter is the only substantive gap for a three-parameter tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, but the description compensates for two of three params by stating the 1–12 track limit and the fixed count of three captions. It never mentions the 'language' parameter or its en/zh/id enum, leaving that enum to the schema alone.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description names a specific verb and artifact ('Prepare a free local-workspace JSON draft') and quantifies scope ('1–12 song/artist title cards and three original captions'), so the agent knows exactly what is produced. It does not explicitly contrast with the close names get_creation_template or search_playlist, so sibling differentiation is left implicit.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It gives a concrete workflow ('Save the project field as JSON, then import it on the linked workspace') and clear exclusions ('does not render a video, fetch audio or provide song lyrics'), which tells the agent when this tool is and is not appropriate. It stops short of naming the replacement tools for those excluded jobs, so the alternative routing is incomplete.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Search four starter tracks by song, artist or mood. This is a dated starter selection, not a live or complete Spotify catalogue. Returns metadata and listening links only.
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | ||
| query | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already cover safety (readOnly, idempotent, non-destructive, closed-world), so the description adds genuine value by disclosing that results are a dated starter set rather than live data and that the return is metadata plus listening links only. It omits pagination or whatever happens when the query is empty, but the core behavioral caveats are stated.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Three short sentences, each earning its place: purpose first, scope caveat second, return format last. Zero filler and front-loaded.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple two-parameter read tool with full annotation coverage and no output schema, the description supplies the essential caveats (dated, non-exhaustive, links only). The remaining gap is the unexplained `limit` parameter and the apparent conflict between the fixed "four tracks" framing and the 1-12 limit range.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, so the description must carry the parameter burden. It clarifies that the query matches across song, artist and mood, which is real added meaning, but leaves `limit` entirely unexplained and creates tension by saying "four starter tracks" while the schema allows limit up to 12.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description opens with a specific verb and resource ("Search four starter tracks by song, artist or mood") and immediately bounds what is searched ("dated starter selection, not a live or complete Spotify catalogue"). Sibling tools (get_creation_template, prepare_video_project) are unrelated in domain, so no explicit differentiation is needed, though none is offered.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It gives clear context for when the tool applies (starter/seed tracks only, matched by song, artist or mood) and an explicit when-not (not for live or complete catalogue lookups). No named alternative is provided for the live-catalogue case, which keeps it below a 5.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v1.0.0get_creation_templateprepare_video_projectsearch_playlistScored across 3 tools
Each tool maps to a distinct step: searching the starter track list, fetching a scene-prompt template, and assembling a JSON project draft. There is mild thematic overlap between search_playlist and prepare_video_project since both concern songs, but the descriptions make clear that one returns metadata and the other builds a file draft.
All three tools follow a consistent snake_case verb_noun convention (search_playlist, get_creation_template, prepare_video_project). Verb choices accurately reflect each action's read-versus-prepare semantics.
Three tools is minimal but matches a narrowly scoped creative-asset helper with a clear three-step pipeline. It is not thin in the problematic sense, since each tool covers a genuinely different stage of the workflow.
The surface is read-and-prepare only: there is no way to retrieve, list, or revise a previously prepared project draft, and no tool exposes the linked workspace or template list. Rendering, audio fetching, and lyric lookup are explicitly out of scope, so the remaining gaps are structural rather than broken.
Video knowledge base for agents: search your library's transcripts, keyframes and on-screen text.
YouTube transcripts, search, channel/playlist listings and upload tracking for AI agents.
Reproducible, IP-clean assets for agents: seven engines, from print designs to voice and video.
Read-only AgentiScript concept search, catalog, authenticity, license, and approved asset discovery.
