Atlassian MCP (Jira, Confluence)

PassAudited by VirusTotal on May 12, 2026.

Overview

Type: OpenClaw Skill Name: atlassian-mcp Version: 1.0.0 The skill bundle is designed to run a Docker container for Atlassian integration, requiring Jira API credentials. The `SKILL.md` and `scripts/run_mcp_atlassian.sh` clearly define this purpose. The script passes sensitive `JIRA_API_TOKEN` as an environment variable to the Docker container but avoids echoing it to stdout, and uses `docker run --rm` to prevent persistence. There is no evidence of prompt injection, data exfiltration from the host system, or other malicious activities within the analyzed files. The behavior is consistent with its stated purpose.

Findings (0)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

ConcernMedium Confidence
ASI02: Tool Misuse and Exploitation
What this means

A mistaken or hijacked agent workflow could create or change Jira work items or projects under your account.

Why it was flagged

This exposes Jira management and mutation workflows to the agent, but the artifacts do not define project limits, read-only operation, human confirmation, or rollback behavior.

Skill content
Reference the container as an MCP source in your Clawdbot config to query issues, create tasks, or manage Jira directly from your agent.
Recommendation

Use least-privilege, project-limited credentials where possible and require explicit user confirmation before create, update, delete, or project-management actions.

What this means

If the remote image changes or is compromised, it could use the supplied Jira token with the permissions of that token.

Why it was flagged

The skill pulls an unpinned remote Docker image and passes a Jira API token into it; the image implementation is not included in the reviewed artifacts.

Skill content
docker pull ghcr.io/sooperset/mcp-atlassian:latest ... JIRA_API_TOKEN=your_api_token ... ghcr.io/sooperset/mcp-atlassian:latest
Recommendation

Pin the Docker image to a trusted version or digest, verify the upstream project, review or build the image yourself if possible, and rotate the token if you later distrust the image.

What this means

Actions taken through the MCP server can operate as the Jira user associated with the token.

Why it was flagged

The script intentionally passes Jira account credentials to the MCP container. This is expected for the integration, but the token may carry broad account permissions.

Skill content
-e JIRA_USERNAME="$JIRA_USERNAME" \
  -e JIRA_API_TOKEN="$JIRA_API_TOKEN" \
Recommendation

Use a dedicated service account or least-privilege token rather than a personal or admin token, and restrict Jira permissions to the projects and actions needed.

What this means

Private Atlassian issue or page content may be visible in the agent session and any downstream outputs the user requests.

Why it was flagged

Jira or Confluence data returned by the MCP tools will be made available to the agent context. This is purpose-aligned, but it is a sensitive data flow users should understand.

Skill content
Once running, the MCP server exposes Jira tools for use.
Recommendation

Connect only workspaces and projects you are comfortable exposing to the agent, and avoid asking the agent to share sensitive ticket or page contents outside approved channels.