fix: Added rh-sre CLAUDE.md

[dmartinol]

Summary

Added rh-sre CLAUDE.md

Pack(s) affected

• rh-sre
• rh-developer
• ocp-admin
• rh-virt
• rh-ai-engineer
• Other / repo-wide

Change type

• New skill
• New agent
• New pack
• Update existing skill / agent
• MCP server config (.mcp.json)
• Docs / README
• CI / tooling

CLAUDE.md compliance

• Agents orchestrate skills; no direct MCP/tool calls in agents
• Skills are single-purpose task executors
• Skills encapsulate all tool access (MCP tools invoked only inside skills)
• Document consultation: file is read with the Read tool, then declared to the user
• No credentials hardcoded; env vars used via ${VAR} references
• Human-in-the-loop confirmation added for any destructive or critical operations

Validation

• make validate passes locally
• New/changed skills have valid YAML frontmatter (name, description)
• New/changed agents have valid YAML frontmatter (name, description)

[github-actions]

Gemini Code Review

1. Summary

This PR introduces the CLAUDE.md instruction router for the rh-sre pack. It defines the SRE assistant persona, enforces the skill-first rule, maps user intents to specific skills, outlines skill chaining/orchestration, documents the required MCP servers, and establishes global safety and execution rules.

2. Project Rules Compliance

This PR perfectly aligns with the project rules, specifically Principle #11 (Pack-Level CLAUDE.md).

• Required Sections: ## Skill-First Rule, ## Intent Routing, ## MCP Servers, and ## Global Rules are all present and correctly formatted.
• Skill Invocation: Correctly enforces the /skill-name slash format (Principle Rh developer/human in the loop #4).
• Security: Explicitly enforces credential safety and ${ENV_VAR} usage (Principle Rh developer/troubleshooting #8).
• Human-in-the-Loop: Correctly mandates confirmation for destructive/broad changes (Principle Rh virt/agent init #6).

Reminders:

• Since this modifies pack configuration, please remember to run make validate locally to ensure the structure passes the automated linter.
• If rh-sre is a newly formalized pack in the build system, ensure it is added to PACK_DIRS in scripts/validate_structure.py. (Ignore if already present, as rh-sre is the reference implementation).

3. Code Quality Issues

No issues found. The Markdown is clean, the table is well-structured, and the instructions are highly specific and actionable for the AI agent.

4. Suggestions

Minor non-blocking thought:
In the ## Skill-First Rule section, you wrote:

"To invoke a skill, use the Skill tool with the skill name (e.g., /cve-impact, /remediation)."

This is generally fine, but to strictly align with Principle #4 ("Avoid - Skill tool format... Use - Slash format"), you might want to phrase it simply as:

"To invoke a skill, simply output the slash command (e.g., /cve-impact, /remediation)."
(Feel free to ignore this if the target marketplace specifically requires invoking a literal tool named Skill with the slash-formatted string as an argument).

5. Verdict

APPROVE — Excellent implementation of the pack-level instruction router.

---

Automated review by Gemini 3.1 Pro | Rules: CLAUDE.md, SKILL_DESIGN_PRINCIPLES.md

[dmartinol]

4. Suggestions

Minor non-blocking thought: In the ## Skill-First Rule section, you wrote:

"To invoke a skill, use the Skill tool with the skill name (e.g., /cve-impact, /remediation)."

This is generally fine, but to strictly align with Principle #4 ("Avoid - Skill tool format... Use - Slash format"), you might want to phrase it simply as:

"To invoke a skill, simply output the slash command (e.g., /cve-impact, /remediation)."
(Feel free to ignore this if the target marketplace specifically requires invoking a literal tool named Skill with the slash-formatted string as an argument).

Suggestion rejected because the given statement matches the Skill-First Rule definition in the other plugins