Single tool 'snapshot_create' has a detailed input schema with 12 parameters, each with type and description. However, the tool description (89 chars) falls below the 10-1024 char baseline and lacks critical context: WHEN to use it, WHAT it returns, and WHY it matters. Parameter descriptions are present but generic (e.g., 'Snapshot profile name' without explaining available profiles or consequences). No output schema is documented, LLMs cannot plan downstream steps. The tool performs a WRITE operation (file creation) but the description does not explicitly state this, violating the command-tool pattern. No error handling guidance is provided.

• Expand tool description to 150-250 chars: 'Create a deterministic snapshot of a repository for analysis. Returns snapshot_path, file_count, total_bytes, and any redaction_warnings. Use this to prepare code for review, archival, or LLM analysis. Respects file size limits and timeout constraints.'
• Document the output schema explicitly: {snapshot_path: string, file_count: int, total_bytes: int, redacted_secrets_count: int, errors: [{path: string, reason: string}], profile_used: string}
• Convert 'profile' parameter to an enum by exposing profile_names from snapshot_profiles.py. Add description: 'Snapshot profile (e.g., minimal, standard, comprehensive). Call list_snapshot_profiles() to discover available options.'
• Add error handling guidance to description: 'Raises RepoGroundMcpToolError if timeout_seconds ≤ 0 or paths are invalid. Raises RepoGroundMcpToolTimeout if execution exceeds timeout. Adjust timeout_seconds (max 1800) or max_file_bytes if limits are exceeded.'
• Add a companion tool 'list_snapshot_profiles' that returns available profiles with descriptions, enabling agents to discover valid options without hardcoding.