feat(agent-production-operations): add agent production operations bundle (#229)

Add the agent-production-operations bundle under bundles/agent-production-operations/.
This bundle provides a runtime control plane for taking an evaluated agent with
tools and authority into controlled production operation.

Deliverables:
- SKILL.md: thin umbrella with runtime control plane routing table composing
  agent-evals-and-observability, release-engineering, site-reliability-engineering,
  secure-software-engineering, and platform-engineering
- references/agent-production-contract.md: capability, authority, uncertainty,
  escalation, and side-effect contracts
- references/runtime-control-plan.md: versioning (model, prompt, tool, policy,
  evaluator), staged rollout (4 stages), and fallback paths
- references/tool-authority-health.md: tool availability/failure and authority
  usage/breach state record
- references/trace-to-eval-feedback.md: production-to-evaluation feedback loop
- references/discovery-brief.md: bounded discovery brief defining boundaries
  with specialist skills
- README.md: human-facing overview with all required sections
- AGENTS.md: agent loading and nested-skill discovery notes
- evals/evals.json: 7 integrated eval cases covering read-only agent, tool-using
  agent with side effects, model regression, tool outage, cost-budget breach,
  human escalation, and incident-learning-driven disablement

Closes #201

Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>

.claude-plugin/marketplace.json

@@ -32,6 +32,15 @@
       "strict": false,
       "description": "Design, run, review, or release framework- and vendor-neutral evaluations and observability for AI agents. Use when defining agent evals, datasets, graders, trajectory review, regression analysis, release gates, production traces, or privacy-aware telemetry. Covers task and trajectory contracts, statistical comparisons, and incident-to-case learning; route framework implementation to pydanticai or langgraph when needed."
     },
+    {
+      "name": "agent-production-operations",
+      "source": "./",
+      "skills": [
+        "./agent-production-operations"
+      ],
+      "strict": false,
+      "description": "Operate an evaluated agent with tools and authority in production through a runtime control plane covering versioning, staged rollout, fallback, cost and latency budgets, tool health, human escalation, disablement, and trace-to-eval feedback. Do not use for building agents, designing evals, or general release, SRE, security, or platform engineering — those methods stay with their specialist skills."
+    },
     {
       "name": "agent-skills",
       "source": "./",