All plugins

marcora

Brand context, marketing blueprints, content creation, grounding, and GTM planning through Marcora's hosted MCP server (Streamable HTTP at https://mcp.marcora.ai/, OAuth 2.1 PKCE or Bearer API token). Bundles the marcora-mcp workflow skill. Agent Plugins v1 package (plugin.json + mcp.json + skills/), configuration only. Disclosure — every tool call goes to Marcora's hosted server (mcp.marcora.ai) over HTTPS and acts on the user's Marcora workspace under their own role; the 63-tool surface (32 read-only, 31 non-read-only, 15 destructiveHint: true) is served at runtime and is not covered by the pinned sha. Non-read-only tools create or update workspace records, send team invitation emails (invite_user), publish share links (create_external_share), or spend team Marcora Studio Credits (create_content with instructions, produce_plan, create_blueprint_draft, create_project with project_brief_instructions, ask_content_assistant, check_content_grounding, trigger_health_audit_scan, convert_markdown_to_word_doc at 2 credits/conversion, and run_workflow). run_workflow starts a server-side automated agent session using Marcora tools and the user's connected third-party apps (allowed_tools is a prompt instruction to the runner, not a separate permission check), workflows that are Active with an On schedule run on Marcora's server-side scheduler independently of the local client, and ask_content_assistant runs an asynchronous background AI assistant that may edit a document in place. The package marks the server untrusted (timeout: 180), so Hermes's host policy prompts before client-initiated calls to tools the server does not mark read-only, though neither Hermes trust nor advisory MCP annotations prevent server-side scheduled workflow runs or background agent actions once started or activated on Marcora.

ccrompToolsv1.2.26Updated 2 days ago

About

Brand context, marketing blueprints, content creation, grounding, and GTM planning through Marcora's hosted MCP server (Streamable HTTP at https://mcp.marcora.ai/, OAuth 2.1 PKCE or Bearer API token). Bundles the marcora-mcp workflow skill. Agent Plugins v1 package (plugin.json + mcp.json + skills/), configuration only. Disclosure — every tool call goes to Marcora's hosted server (mcp.marcora.ai) over HTTPS and acts on the user's Marcora workspace under their own role; the 63-tool surface (32 read-only, 31 non-read-only, 15 destructiveHint: true) is served at runtime and is not covered by the pinned sha. Non-read-only tools create or update workspace records, send team invitation emails (invite_user), publish share links (create_external_share), or spend team Marcora Studio Credits (create_content with instructions, produce_plan, create_blueprint_draft, create_project with project_brief_instructions, ask_content_assistant, check_content_grounding, trigger_health_audit_scan, convert_markdown_to_word_doc at 2 credits/conversion, and run_workflow). run_workflow starts a server-side automated agent session using Marcora tools and the user's connected third-party apps (allowed_tools is a prompt instruction to the runner, not a separate permission check), workflows that are Active with an On schedule run on Marcora's server-side scheduler independently of the local client, and ask_content_assistant runs an asynchronous background AI assistant that may edit a document in place. The package marks the server untrusted (timeout: 180), so Hermes's host policy prompts before client-initiated calls to tools the server does not mark read-only, though neither Hermes trust nor advisory MCP annotations prevent server-side scheduled workflow runs or background agent actions once started or activated on Marcora.

Marcora MCP Server

Marcora is the context layer for go-to-market teams: your positioning, brand voice, and product facts in one source of truth that your AI tools read from.

Learn more: marcora.ai · Docs: marcora.ai/docs/mcp-overview

Add to Kiro

About

This is the public documentation and plugin repository for Marcora's hosted MCP server. The server itself is a commercial, closed-source remote service. This repo provides Portable Agent Plugin v1 metadata (plugin.json + mcp.json), the companion marcora-mcp workflow skill, connection instructions, tool documentation, and support resources.

TypeHosted remote MCP server
ImplementationClosed-source (commercial)
TransportStreamable HTTP
AuthenticationOAuth 2.1 (PKCE) or API token
StatusProduction

What It Does

Marcora is the context layer for go-to-market teams. Build your source of truth once: a Brand Foundation (positioning, brand voice, and writing style) plus a Reference Library of product docs, buyer personas, competitive analysis, and messaging. Claude, ChatGPT, and any AI tool that supports MCP connections read from it, so what you write stays on-brand and accurate.

Marcora was formerly called MarketCore.

Marcora's MCP server enables AI clients to:

  • Manage brand and company context and reference materials
  • Create and manage blueprints for content generation
  • Generate marketing content from context and blueprints
  • Organize work with projects and collections
  • Browse and import community blueprints and templates
  • Plan content on the plans board and produce it, individually or from reusable playbooks
  • Automate recurring work with reusable, multi-step workflows
  • Share content externally and export to Word

Connection

Server URL (OAuth): https://mcp.marcora.ai/

Authentication options:

  • OAuth 2.1 with PKCE (recommended for interactive clients) — connect using https://mcp.marcora.ai/
  • API Token (for non-interactive environments) — generate a key at Integration Settings and connect to the same URL (https://mcp.marcora.ai/), passing the key in an Authorization: Bearer <your-api-key> header. The server negotiates the right transport automatically.

Quickstart

Add to your MCP client config (mcp.json / .kiro/settings/mcp.json / claude_desktop_config.json):

{
  "mcpServers": {
    "marcora": {
      "url": "https://mcp.marcora.ai/"
    }
  }
}
  • Kiro: Click Add to Kiro above, or add the snippet to ~/.kiro/settings/mcp.json (user) or .kiro/settings/mcp.json (workspace). Kiro signs in via browser OAuth on first connection.
  • Hermes Agent: Run hermes plugins install marcora (or hermes plugins install https://github.com/ccromp/marcora-mcp) to install both the remote MCP server connection and the bundled marcora-mcp workflow skill.

See docs/quickstart.md for setup instructions for Claude, ChatGPT, Claude Desktop, Cursor, Kiro, Hermes Agent, and VS Code.

The Marcora AI Workflows Skill (recommended)

The Marcora AI Workflows Anthropic Skill ships alongside this server. It teaches your AI client Marcora's mental model, the right tool sequencing, the questions to ask the user before acting, and the pitfalls to avoid — so the agent stops misfiring on tasks that look ambiguous from tool descriptions alone.

SkillWhat it coversSource
Marcora AI Workflows (marcora-mcp)Object model (content, blueprints, projects, context items); content-generation patterns; the 4-layer context model; project-brief mechanics; the content plans board and playbooks; building reusable multi-step workflows; and choosing between similar toolsskill/marcora-mcp/

Your AI client picks the skill up on demand based on the task.

Install with the skills CLI

npx skills add ccromp/marcora-mcp

This installs the marcora-mcp skill for Claude Code and other agents that read SKILL.md skills. To preview first, run npx skills add ccromp/marcora-mcp --list.

Download

Get the .skill file from the latest release:

  • marcora-mcp.skill — the Marcora skill (a zip archive containing SKILL.md and its bundled reference files).

Install

Claude Code:

# Download and unzip into your skills directory
mkdir -p ~/.claude/skills
unzip ~/Downloads/marcora-mcp.skill -d ~/.claude/skills/

Restart Claude Code. The skill loads automatically on any matching task.

Claude Desktop:

Import the .skill file via Claude Desktop's Skills folder. See Anthropic's Skills documentation for the current path on your platform.

ChatGPT, Cursor, and other non-Anthropic clients:

These clients don't directly support Anthropic Skills (yet). Workaround: unzip the .skill file and paste the contents of SKILL.md into your client's custom instructions or system-prompt slot. Bundled reference files can be pasted in too if your client supports longer context.

Power-user install (clone and symlink)

If you'd rather track main directly:

git clone https://github.com/ccromp/marcora-mcp.git
ln -s "$(pwd)/marcora-mcp/skill/marcora-mcp" ~/.claude/skills/marcora-mcp

The unbundled skill source lives under skill/.

Releases

Releases are tagged marcora-vX.Y.Z — see Releases. Each release attaches the bundled skill as marcora-mcp.skill. The skill's own version is the metadata.version field in its SKILL.md, and it moves independently of the release tag.

Available Tools

CategoryToolDescription
Accountget_current_user_infoGet profile, subscription, and usage info
Accountget_team_infoList every team you belong to + each team's members (numeric user_ids for assignment)
Accountset_active_teamSwitch your active team (returns previous_team_id to restore)
Context & Resourcesget_brand_foundationGet team's Brand Foundation (company overview, voice, style, examples)
Context & Resourcesupdate_brand_foundationOverwrite one Brand Foundation element
Context & Resourceslist_context_collectionsList context collections
Context & Resourcescreate_context_collectionCreate a new context collection
Context & Resourcesadd_contextAdd a context item to your library
Context & Resourcesupdate_contextUpdate an existing context item
Context & Resourceslist_context_itemsList context items in your library (optional semantic search)
Context & Resourcesget_context_itemGet a context item's full markdown content
Context & Resourcesget_relevant_contextSearch context library by prompt (with a project_id, also returns the project's brief)
Context Intelligencelist_ci_findingsList findings from automated context-library scans
Context Intelligenceget_ci_findingGet one finding in full detail by UUID
Context Intelligenceupdate_ci_finding_statusAcknowledge, dismiss, or resolve a finding
Context Intelligencetrigger_health_audit_scanStart a library-wide context health audit
Content Groundingcheck_content_groundingCheck a document's claims against your context library
Content Groundingget_grounding_resultRead a grounding scan's result (poll with scan_id)
Content Groundingapply_grounding_fixApply a finding's recommended fix by id
Referencelist_content_categoriesList content categories
Referencelist_targeting_dimensionsList targeting dimensions and options
Referencelearn_marcora_toolsMarcora's guide to its own tools, by topic (for assistants without the skill)
Blueprintslist_blueprintsList all blueprints
Blueprintsget_blueprintGet blueprint details by UUID
Blueprintscreate_blueprintCreate a reusable content template
Blueprintscreate_blueprint_draftCreate an AI-assisted blueprint draft
Blueprintsfinalize_blueprint_draftPublish a blueprint draft
Community Blueprintslist_community_blueprintsBrowse community templates
Community Blueprintsget_community_blueprint_detailsGet community blueprint details
Community Blueprintsimport_community_blueprintImport a community blueprint
Contentcreate_contentSave a new document in the workspace: your own text by default, or written by Marcora (with or without blueprint)
Contentget_generation_statusCheck async generation status
Contentlist_contentList all content (optional semantic search)
Contentget_contentGet full content by ID
Contentupdate_contentUpdate a content document
Contentask_content_assistantRun the in-document Content Assistant on a document
Sharing & Exportcreate_external_shareCreate a public share link
Sharing & Exportget_share_link_activityCheck whether a share link was opened, by how many viewers, and when
Sharing & Exportlist_share_link_activityList the workspace's share links with their activity
Sharing & Exportconvert_markdown_to_word_docExport markdown as Word doc
Projectslist_projectsList all projects
Projectsget_projectGet project details, including its brief
Projectscreate_projectCreate a new project
Projectsupdate_projectUpdate mutable fields on a project
Plans & Playbookslist_plansList content plans (with filters)
Plans & Playbooksget_planGet plan details by UUID
Plans & Playbookscreate_planCreate a new content plan
Plans & Playbooksupdate_planUpdate a plan or transition its stage
Plans & Playbooksproduce_planProduce (generate) the content a plan describes
Plans & Playbookslist_playbooksList content playbooks
Plans & Playbooksget_playbookGet a playbook and its ordered items
Plans & Playbookscreate_playbookAuthor a reusable playbook of plan items
Plans & Playbookscreate_playbook_from_plansCapture existing plans as a playbook
Plans & Playbooksupdate_playbookRename, re-describe, or restructure a playbook
Plans & Playbooksinstantiate_playbookRun a playbook into a batch of plans (optionally anchored to a date)
Workflowslist_workflowsList workflows
Workflowsget_workflowGet a workflow's full definition
Workflowscreate_workflowCreate a new workflow template
Workflowsupdate_workflowUpdate, activate, or archive a workflow
Workflowsrun_workflowRun a workflow now
Workflowsget_workflow_runsInspect workflow run history

See docs/tools.md for the complete tool reference.

Documentation

Security, Privacy & Runtime Disclosures

  • External service & authentication: This package connects over HTTPS to Marcora's hosted MCP server at https://mcp.marcora.ai/. Connecting requires a Marcora account (a permanent Free tier is available at marcora.ai/pricing) authenticated via OAuth 2.1 with PKCE (/oauth/authorize, /oauth/token, /oauth/register, /oauth/revoke) or a user-generated Bearer API token from Integration Settings.
  • Runtime tool surface (63 tools) & Studio Credit costs: At runtime, https://mcp.marcora.ai/ serves 63 MCP tools: 32 read-only tools (readOnlyHint: true, destructiveHint: false, always free to call) and 31 non-read-only tools (readOnlyHint: false, of which 16 have destructiveHint: false and 15 have destructiveHint: true because they overwrite existing records in place, trigger scans/workflows, or send team invites). Those 31 tools create or update workspace records or consume team Marcora Studio Credits when AI generation, scans, or Word exports run (create_content on the instructions generation path, produce_plan, create_blueprint_draft, create_project with project_brief_instructions, ask_content_assistant, check_content_grounding, trigger_health_audit_scan, convert_markdown_to_word_doc at 2 Studio Credits per conversion, and run_workflow).
  • Server-side background agents, scheduled workflows & connected apps: Calling run_workflow starts an automated agent session on Marcora's backend that executes the workflow's steps using Marcora tools and any third-party apps the user has connected in Marcora (allowed_tools on a workflow is a prompt instruction to the workflow runner, not a separate hard permission check). Workflows created (create_workflow) or activated (update_workflow with status: "active") that have an On schedule are dispatched automatically by Marcora's server-side scheduler independently of the local MCP client. Calling ask_content_assistant starts an asynchronous background AI assistant run on an existing document that may edit the document body in place.
  • Advisory nature of MCP annotations & Hermes trust: "untrusted" scope: MCP ToolAnnotations (readOnlyHint, destructiveHint, idempotentHint, openWorldHint) are advisory hints; each MCP host decides whether and how to prompt for confirmation. In Hermes Agent, plugin.json declares "trust": "untrusted" and "timeout": 180 on the marcora server so Hermes's host policy prompts before client-initiated calls to non-read-only tools, but client-side trust: "untrusted" and MCP annotations only govern calls initiated by the local client and do not prevent server-side scheduled workflow runs or background agent execution once a workflow is started or activated on Marcora.
  • Privacy Policy: marcora.ai/privacy-policy
  • Terms of Service: marcora.ai/terms
  • Security & Trust Center: marcora.ai/security and docs/security.md

Support

License

The documentation, skill, and metadata in this repository are licensed under the MIT License. The Marcora MCP server implementation is proprietary and is not covered by that license or included in this repository.

README from the reviewed commit 41d1c6d; it updates when the author re-pins.

Developer Mode

Explore built-in Hermes toolsets and actions.

HermesAgent

Developer?

Create a Hermes extension for Desktop or agent workflows.

Developer Section