The server provides 31 tools with partial documentation. All tools have names following verb_noun conventions (good naming), and descriptions are present for each. However, schema documentation is incomplete and inconsistent. Input parameters are documented with descriptions and type information, but output schemas are not explicitly documented anywhere in the provided code. Parameter descriptions vary in quality, most are adequate but generic. Several critical gaps: no parameter constraints (enums for risk levels, project keys, etc.), no clear error handling guidance, no indication of pagination support despite tools returning lists, and no explicit acknowledgment of destructive vs safe operations in descriptions. The server handles multiple Atlassian products (Jira, Confluence, Bitbucket) with reasonable coverage, but consistency and depth of documentation would benefit production deployment.

• Document output schema for every tool in the form: 'Returns: {field_name: type, ...}'. For example, jira_get_issue should document: 'Returns: {issueKey: string, summary: string, description: string, status: string, assignee: {name: string, email: string}, ...}'.
• Convert free-form string parameters to enums where values are known and finite. E.g., issueType in jira_create_issue should be enum: [Bug, Story, Task, ...]. Update descriptions to list valid values explicitly.
• Add explicit idempotency disclaimers to all write/create operations. E.g., 'Note: This operation is not idempotent. Retrying with the same parameters may create duplicates.' Or offer an optional idempotencyKey parameter.
• For destructive operations (delete_*), add a confirmation step or require explicit confirmation parameter. E.g., 'jira_delete_issue requires a confirmation=true parameter and should return a confirmation token before permanent deletion.'
• Enhance parameter descriptions with format constraints and examples (without treating examples as literal defaults). E.g., 'issueKey: The Jira issue identifier, formatted as PROJECT-NUMBER (e.g., PROJ-42, where PROJECT is 2-10 uppercase letters).'
• Document pagination for all list-returning tools. E.g., 'Returns results in pages of up to maxResults items. Use the startIndex field from the response to fetch the next page. If totalResults <= startIndex + maxResults, no more pages exist.'
• Add error recovery hints to tool descriptions. E.g., 'If the issue is not found, try jira_search_issues() to locate it by summary or JQL query.' or 'If authentication fails, verify your Atlassian API token is configured correctly.'

Inferred effective spec: <=2025-11-25.

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry