#16087: docs: enhance OpenCode documentation in coding-agent skill

## Summary - Enhanced OpenCode CLI documentation in `skills/coding-agent/SKILL.md` to match the detail level of Codex CLI and Claude Code sections - Added comprehensive command reference table (run, attach, session, pr, models, mcp) - Documented all `opencode run` flags with descriptions - Added practical examples for session management, multi-model support, and GitHub PR integration ## Changes The original OpenCode section was minimal (5 lines). This PR expands it to include: 1. **Command Reference**: All major OpenCode CLI commands 2. **Flags Table**: Complete `opencode run` flag documentation 3. **Basic Usage**: One-shot and background execution examples 4. **Session Management**: Continue, fork, and list sessions 5. **Multi-Model Support**: Examples with different providers (Anthropic, OpenAI, Gemini, DeepSeek, OpenCode Zen) 6. **GitHub PR Integration**: `opencode pr` command usage 7. **MCP Section**: Reference to MCP server management ## Testing - Verified markdown formatting is correct - Confirmed all command examples match OpenCode CLI help output --- **Note**: This PR was created with AI assistance. The changes are documentation-only and follow the existing patterns in the file. <!-- greptile_comment --> <h3>Greptile Summary</h3> This documentation-only PR expands the OpenCode section in `skills/coding-agent/SKILL.md` from a minimal 2-line example to comprehensive documentation matching the detail level of the Codex CLI section. The additions include a command reference table, `opencode run` flags table, and practical examples for session management, multi-model support, GitHub PR integration, and MCP server management. - All examples follow the established `bash pty:true` pattern used throughout the file - The `opencode/claude-opus-4-6` model reference in the Multi-Model section is consistent with the codebase (`src/commands/opencode-zen-model-default.ts`) - The new subsections (Commands, Flags, Basic Usage, Session Management, GitHub PR Integration, Multi-Model Support, MCP) mirror the structure of the Codex CLI section <h3>Confidence Score: 5/5</h3> - This PR is safe to merge — it contains only documentation additions to a skill markdown file with no code changes. - Documentation-only change that adds well-structured content following existing patterns in the file. No code logic, no configuration changes, no risk of runtime issues. The model references are consistent with the codebase. - No files require special attention. _{Last reviewed commit: 5a2f9e5} <!-- greptile_other_comments_section --> _{(3/5) Reply to the agent's comments like "Can you suggest a fix for this @greptileai?" or ask follow-up questions!} <!-- /greptile_comment -->