← Back to Skills Marketplace
maverick

Confluence mcp

by Maverick · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ Security Clean
41
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install maverick-confluence-mcp
Description
Search, read, and update Confluence pages, spaces, comments, and documentation workflows via Atlassian's hosted MCP server (https://mcp.atlassian.com/v1/mcp/...
README (SKILL.md)

Confluence

Quick start

Always invoke through bash {baseDir}/scripts/invoke.sh — never call mcporter directly. The wrapper seeds the OAuth vault from the env-supplied tokens when needed, then calls mcporter.

bash {baseDir}/scripts/invoke.sh list maverick-confluence --schema

For structured output (also surfaces transport errors as JSON envelopes — workaround for mcporter #153):

bash {baseDir}/scripts/invoke.sh call --output json maverick-confluence.TOOL_NAME key=value | jq '.result.content'

Safety

Write operations that create, edit, publish, archive, delete, move, comment on, or bulk-update pages modify Confluence content visible to the connected Atlassian workspace. Confirm clear user intent before invoking write tools — search and read tools are safe to call freely while exploring. Search spaces and pages before assuming IDs, and read current content plus version metadata before editing.

Authentication

Tokens are provisioned and rotated automatically. If a call returns HTTP 401 that doesn't recover within a few seconds, the OAuth grant has been revoked — re-authorize the integration to refresh credentials.

Data flow

Tool calls travel to Atlassian's hosted MCP service at https://mcp.atlassian.com/v1/mcp/authv2 over HTTPS, authenticated via OAuth. Atlassian sees the Confluence page, space, comment, and workspace data referenced by each call. Use this skill for Confluence-related work only; do not pass unrelated sensitive content through these tools.

Dependencies

  • mcporter (github.com/steipete/mcporter) — MCP CLI used to invoke Atlassian's hosted MCP server. Auto-installed via npm install -g --ignore-scripts mcporter if missing on PATH (see install spec in frontmatter). The install spec uses unpinned mcporter (npm latest); operators with strict supply-chain controls should override the install to pin a specific version (e.g. mcporter@\x3Cversion>).
  • jq (stedolan.github.io/jq) — JSON processor used by the vault initializer. System dependency; install via your OS package manager (apt install jq, brew install jq, etc.).
  • flock (part of util-linux) — file locking used to serialize concurrent vault writes. Available by default on Linux; on macOS install via brew install flock.
  • shasum (Perl, ships with Digest::SHA) — computes the SHA-256 hashes used to derive the mcporter vault key and the provisioned-token marker. Preinstalled on macOS and on Debian/Ubuntu (incl. the deployed cloudflare/sandbox Ubuntu 22.04 image); on minimal Linux images install perl-Digest-SHA. The script invokes shasum -a 256 rather than GNU sha256sum so it runs on stock macOS without coreutils.
Usage Guidance
This appears to be a legitimate Confluence MCP integration. Before installing, make sure the OAuth grant has only the Confluence permissions you intend, protect the local mcporter credential vault, pin mcporter if your environment requires supply-chain control, and personally review any action that edits, deletes, archives, moves, comments on, publishes, or bulk-updates Confluence content.
Capability Analysis
Type: OpenClaw Skill Name: maverick-confluence-mcp Version: 1.0.0 The skill provides a legitimate integration with Atlassian Confluence via the Model Context Protocol (MCP). It uses the 'mcporter' tool to communicate with Atlassian's official MCP server (mcp.atlassian.com). The scripts 'init-mcporter.sh' and 'invoke.sh' are designed to securely seed OAuth tokens from environment variables into the local mcporter vault, employing security best practices such as passing secrets to jq via environment variables to avoid exposure in process lists and using file locking for concurrency. No evidence of malicious intent, data exfiltration to unauthorized endpoints, or harmful prompt injection was found.
Capability Tags
requires-oauth-tokenrequires-sensitive-credentials
Capability Assessment
Purpose & Capability
The skill’s ability to search, read, and update Confluence pages, spaces, and comments matches its stated purpose, but write/delete/bulk-update operations can materially affect a shared workspace.
Instruction Scope
The instructions are scoped to Confluence work and explicitly tell the agent to confirm clear user intent before write operations.
Install Mechanism
The install path uses the npm package mcporter without a pinned version; this is disclosed and purpose-aligned, but operators may want to pin or verify the dependency.
Credentials
The required OAuth tokens and Atlassian-hosted MCP data flow are expected for a Confluence integration and are disclosed, but they grant access to workspace data.
Persistence & Privilege
The wrapper seeds persistent OAuth state into ~/.mcporter/credentials.json and stores a local hash marker in the skill directory; this is explained in the script and appears scoped to mcporter use.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install maverick-confluence-mcp
  3. After installation, invoke the skill by name or use /maverick-confluence-mcp
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of maverick-confluence-mcp. - Search, read, and update Confluence pages, spaces, comments, and documentation workflows via Atlassian's hosted MCP server. - Uses OAuth-based authentication with automatic token provisioning and rotation. - Provides structured output and error handling through a wrapper script. - Includes safety guidelines for write operations to Confluence content. - Auto-installs required dependencies: mcporter, jq, flock, and shasum.
Metadata
Slug maverick-confluence-mcp
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Confluence mcp?

Search, read, and update Confluence pages, spaces, comments, and documentation workflows via Atlassian's hosted MCP server (https://mcp.atlassian.com/v1/mcp/... It is an AI Agent Skill for Claude Code / OpenClaw, with 41 downloads so far.

How do I install Confluence mcp?

Run "/install maverick-confluence-mcp" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Confluence mcp free?

Yes, Confluence mcp is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Confluence mcp support?

Confluence mcp is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Confluence mcp?

It is built and maintained by Maverick (@maverick); the current version is v1.0.0.

💬 Comments