prompts.chat MCP Server

Measured conclusion: Deep-tested. L2 outcome: tools_ok.

Add to your agent

Copy this into your MCP client config:

{
  "mcpServers": {
    "prompts.chat MCP Server": {
      "url": "https://prompts.chat/api/mcp",
      "transport": "streamable-http"
    }
  }
}

Measured tools (10)

• search_prompts

  Search for AI prompts by keyword. Returns matching prompts with title, description, content, author, category, and tags. Use this to discover prompts for various AI tasks like coding, writing, analysis, and more.

  Parameters:

  • query (required) — string — Search query to find relevant prompts
  • tag — string — Filter by tag slug
  • type — string — Filter by prompt type
  • limit — number — Maximum number of prompts to return (default 10, max 50)
  • category — string — Filter by category slug

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• get_prompt

  Get a prompt by ID and optionally fill in its variables. If the prompt contains template variables (like {{variable}}), you will be asked to provide values for them.

  Parameters:

  • id (required) — string — The ID of the prompt to retrieve
  • fill_variables — boolean — If true and the prompt has template variables, triggers interactive variable filling. Default false — returns raw prompt with variable metadata.

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• save_prompt

  Save a new prompt to your prompts.chat account. Requires API key authentication. Prompts are private by default unless configured otherwise in settings.

  Parameters:

  • title (required) — string — Title of the prompt
  • content (required) — string — The prompt content. Can include variables like ${variable} or ${variable:default}
  • tags — array — Optional array of tag names (will be created if they don't exist)
  • type — string — Prompt type (default: TEXT)
  • category — string — Optional category slug
  • isPrivate — boolean — Whether the prompt is private (default: uses your account setting)
  • description — string — Optional description of the prompt
  • structuredFormat — string — Format for structured prompts

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• improve_prompt

  Transform a basic prompt into a well-structured, comprehensive prompt using AI. Optionally searches for similar prompts for inspiration. Supports different output types (text, image, video, sound) and formats (text, JSON, YAML).

  Parameters:

  • prompt (required) — string — The prompt to improve
  • outputType — string — Content type: text, image, video, or sound
  • outputFormat — string — Response format: text, structured_json, or structured_yaml

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• save_skill

  Save a new Agent Skill to your prompts.chat account. Skills are multi-file prompts that can include SKILL.md (required), reference docs, scripts, and configuration files. Requires API key authentication. If the file contents are too long, first save the SKILL.md only, with no other files. Then call add_file_to_skill tool per file.

  Parameters:

  • title (required) — string — Title of the skill
  • files (required) — array — Array of files. Must include SKILL.md as the main skill file.
  • tags — array — Optional array of tag names
  • category — string — Optional category slug
  • isPrivate — boolean — Whether the skill is private (default: uses your account setting)
  • description — string — Description of what the skill does

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• add_file_to_skill

  Add a new file to an existing Agent Skill. Use this to add reference docs, scripts, or configuration files to a skill you own.

  Parameters:

  • skillId (required) — string — The ID of the skill to add the file to
  • filename (required) — string — File path (e.g., 'reference.md', 'scripts/helper.py', 'config/settings.json')
  • content (required) — string — File content

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• update_skill_file

  Update an existing file in an Agent Skill. Use this to modify reference docs, scripts, configuration files, or SKILL.md content.

  Parameters:

  • skillId (required) — string — The ID of the skill containing the file
  • filename (required) — string — File path to update (e.g., 'SKILL.md', 'reference.md', 'scripts/helper.py')
  • content (required) — string — New file content

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• remove_file_from_skill

  Remove a file from an existing Agent Skill. Cannot remove SKILL.md as it is required.

  Parameters:

  • skillId (required) — string — The ID of the skill to remove the file from
  • filename (required) — string — File path to remove (e.g., 'reference.md', 'scripts/helper.py')

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• get_skill

  Get an Agent Skill by ID, including all its files (SKILL.md, reference docs, scripts, etc.). Returns the skill metadata and file contents. Save to .claude/skills/{slug}/SKILL.md and .claude/skills/{slug}/[other files] structure if user asks to download.

  Parameters:

  • id (required) — string — The ID of the skill to retrieve

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• search_skills

  Search for Agent Skills by keyword. Returns matching skills with title, description, author, and file list. Use this to discover reusable AI agent capabilities for coding, analysis, automation, and more.

  Parameters:

  • query (required) — string — Search query to find relevant skills
  • tag — string — Filter by tag slug
  • limit — number — Maximum number of skills to return (default 10, max 50)
  • category — string — Filter by category slug

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

Measured by MCPYP probing. Untested fields are omitted, never estimated. Listing is always free; MCPYP never proxies calls or stores user tokens.