Atlassian Integration (Jira + Confluence)

Goal

Use Jira and Confluence as external context sources in MCP workflows:

• Jira: search/read issue context
• Confluence: search/read documentation context
• Both are read-focused and audited

What You Need

• Atlassian Cloud site URL (example: https://your-org.atlassian.net)
• Atlassian account email
• Atlassian API token
  • Generate from Atlassian account security settings
• workspace_key in memory-core (example: personal)

Environment Variables (Fallback)

• Jira
  • MEMORY_CORE_JIRA_BASE_URL
  • MEMORY_CORE_JIRA_EMAIL
  • MEMORY_CORE_JIRA_API_TOKEN
• Confluence
  • MEMORY_CORE_CONFLUENCE_BASE_URL
  • MEMORY_CORE_CONFLUENCE_EMAIL
  • MEMORY_CORE_CONFLUENCE_API_TOKEN

Step-by-Step Setup

1. Generate Atlassian API token

• Create one token and reuse it for Jira + Confluence.
• Keep it in your secret manager.

2. Save config in Admin UI

• Open admin-ui -> Integrations.
• Save Jira:
  • enabled=true
  • base_url
  • email
  • api_token
• Save Confluence:
  • enabled=true
  • base_url (https://your-org.atlassian.net or https://your-org.atlassian.net/wiki)
  • email
  • api_token
• Optional:
  • write_on_commit
  • write_on_merge
  • These currently drive audit/hook routing, not provider-side write.

3. Save config via API (optional)

curl -X PUT "$MEMORY_CORE_URL/v1/integrations" \
  -H "Authorization: Bearer $MEMORY_CORE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "workspace_key": "personal",
    "provider": "jira",
    "enabled": true,
    "reason": "enable jira context for team",
    "config": {
      "base_url": "https://your-org.atlassian.net",
      "email": "you@company.com",
      "api_token": "atlassian-token"
    }
  }'

curl -X PUT "$MEMORY_CORE_URL/v1/integrations" \
  -H "Authorization: Bearer $MEMORY_CORE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "workspace_key": "personal",
    "provider": "confluence",
    "enabled": true,
    "reason": "enable confluence context for docs",
    "config": {
      "base_url": "https://your-org.atlassian.net/wiki",
      "email": "you@company.com",
      "api_token": "atlassian-token"
    }
  }'

4. Validate with API

curl -G "$MEMORY_CORE_URL/v1/jira/search" \
  -H "Authorization: Bearer $MEMORY_CORE_API_KEY" \
  --data-urlencode "workspace_key=personal" \
  --data-urlencode "q=deployment incident" \
  --data-urlencode "limit=5"

curl -G "$MEMORY_CORE_URL/v1/confluence/search" \
  -H "Authorization: Bearer $MEMORY_CORE_API_KEY" \
  --data-urlencode "workspace_key=personal" \
  --data-urlencode "q=runbook" \
  --data-urlencode "limit=5"

5. Validate from MCP tools

• jira_search({ q, limit? })
• jira_read({ issue_key, max_chars? })
• confluence_search({ q, limit? })
• confluence_read({ page_id, max_chars? })

API Endpoints

Jira:

• GET /v1/jira/search?workspace_key=<ws>&q=<query>&limit=10
• GET /v1/jira/read?workspace_key=<ws>&issue_key=<ABC-123>&max_chars=4000

Confluence:

• GET /v1/confluence/search?workspace_key=<ws>&q=<query>&limit=10
• GET /v1/confluence/read?workspace_key=<ws>&page_id=<id-or-url>&max_chars=4000

Permissions and Audit

• Read/search requires workspace member access
• All calls are logged to audit_logs:
  • jira.search
  • jira.read
  • confluence.search
  • confluence.read

Env vs Admin UI Priority

• Default: workspace config in Admin UI wins over env fallback.
• Lock option:
  • MEMORY_CORE_INTEGRATION_LOCKED_PROVIDERS=jira,confluence
  • When locked, Admin UI updates are blocked and env-only is enforced.

Troubleshooting

• Invalid API key
  • Check Authorization: Bearer <key>.
• Integration not configured style errors
  • Verify base_url, email, api_token were saved for the workspace.
• Search works but read fails
  • Confirm issue key/page id exists and permission is granted in Atlassian.