Get Skippr Issue Details

Instructions

Implementation Reference

• src/tools/get-issue.ts:43-61 (handler)

  Main handler function for the skippr_get_issue tool. Validates input (projectId, reviewId, issueId), reads the issue markdown file from .skippr/projects/{projectId}/reviews/{reviewId}/issues/{issueId}.md, parses frontmatter, and returns full issue details including raw markdown content.

  export async function getIssue(input: GetIssueInput): Promise<GetIssueOutput> {
    // Validate input
    const validated = GetIssueInputSchema.parse(input);
    const { projectId, reviewId, issueId } = validated;
  
    // Read and parse the issue file
    const { frontmatter, markdown } = await readIssueFile(projectId, reviewId, issueId);
  
    return {
      id: frontmatter.id,
      reviewId: frontmatter.reviewId,
      title: frontmatter.title,
      severity: frontmatter.severity,
      resolved: frontmatter.resolved,
      category: frontmatter.category,
      elementMetadata: frontmatter.elementMetadata,
      markdown, // Raw markdown for Claude to read
    };
  }

• src/tools/get-issue.ts:10-14 (schema)

  Input schema for skippr_get_issue, requiring projectId (string), reviewId (UUID), and issueId (UUID).

  export const GetIssueInputSchema = z.object({
    projectId: z.string().describe('Project identifier'),
    reviewId: z.string().uuid().describe('Review ID'),
    issueId: z.string().uuid().describe('Issue ID'),
  });

• src/tools/get-issue.ts:17-31 (schema)

  Output schema for skippr_get_issue, defining the returned issue details including id, reviewId, title, severity, resolved status, optional category and elementMetadata, plus raw markdown content.

  export const GetIssueOutputSchema = z.object({
    id: z.string().uuid(),
    reviewId: z.string().uuid(),
    title: z.string(),
    severity: z.string(),
    resolved: z.boolean(),
    category: z.string().optional(),
    elementMetadata: z
      .object({
        selector: z.string(),
        bounding_box: z.array(z.number()).optional(),
      })
      .optional(),
    markdown: z.string().describe('Raw markdown content for Claude to read'),
  });

• src/servers/mcp.ts:53-73 (registration)

  Registration of the skippr_get_issue tool in the MCP server using mcpServer.registerTool(). Maps to the getIssue handler and passes its input/output schemas.

  mcpServer.registerTool(
    'skippr_get_issue',
    {
      title: 'Get Skippr Issue Details',
      description: 'Gets full details for a specific Skippr issue including raw markdown content',
      inputSchema: GetIssueInputSchema.shape,
      outputSchema: z.object({
        id: z.string(),
        reviewId: z.string(),
        title: z.string(),
        severity: z.string(),
        resolved: z.boolean(),
        category: z.string().optional(),
        markdown: z.string()
      }).shape
    },
    async (args) => {
      const result = await getIssue(args as GetIssueInput);
      return createStructuredResponse(result);
    }
  );

• src/utils/issues-reader.ts:60-74 (helper)

  Helper that reads the issue markdown file from the filesystem at .skippr/projects/{projectId}/reviews/{reviewId}/issues/{issueId}.md and returns parsed frontmatter plus raw markdown.

  export async function readIssueFile(
    projectId: string,
    reviewId: string,
    issueId: string
  ): Promise<{ frontmatter: ParsedIssueFrontmatter; markdown: string }> {
    const filePath = join(getWorkingDirectory(), '.skippr', 'projects', projectId, 'reviews', reviewId, 'issues', `${issueId}.md`);
  
    const content = await readFile(filePath, 'utf-8');
    const parsed = parseIssueFrontmatter(content);
  
    return {
      frontmatter: parsed.frontmatter,
      markdown: parsed.content, // Raw markdown for Claude to read
    };
  }