Updates

Change History

2026-09-26

Change Summary

• Added Claude Opus 5.5 (released 2026-09-22, $4/$20, $0.20 cache reads = 5% of input) to the current Anthropic lineup table in cost-latency.mdx as the official starting point for most workloads, and relabeled Opus 5 ($5/$25) as the previous default.
• Noted that Sonnet 5's $2/$10 is now the standard price; the scheduled September 1 increase to $3/$15 was cancelled.

Verification Scope

Only the Anthropic rows were rechecked against the Anthropic pricing page. The 2026-05-17 snapshot (including OpenAI and DeepSeek) stays as the budget basis of its own date.

Sources

• Anthropic pricing: https://platform.claude.com/docs/en/about-claude/pricing
• Claude Opus 5.5 overview: https://platform.claude.com/docs/en/models/opus-5-5/overview

2026-09-05

Change Summary

• Added a current Anthropic lineup table to cost-latency.mdx, checked 2026-09-05: Fable 5.1 ($10/$50 with $0.25 cache reads), Opus 5 ($5/$25), Sonnet 5 ($2/$10), and Haiku 4.5 ($1/$5).
• Kept the 2026-05-17 snapshot as the budget basis of its own date and pointed new routing policies at the current table instead.
• Noted that Fable 5.1 charges 2.5% of the input price for cache reads, versus 10% on other Claude models, which changes the cache-reuse economics of long agentic sessions.

Verification Scope

Only the Anthropic pricing and lineup rows were rechecked. The OpenAI and DeepSeek rows and the rest of the chapters remain at their 2026-06-13 verification state.

Sources

• Claude models overview: https://platform.claude.com/docs/en/about-claude/models/overview
• Claude Fable 5.1 overview: https://platform.claude.com/docs/en/models/fable-5-1/overview

2026-06-13

Change Summary

• Normalized a non-standard Callout type in the "Security First" note (warning → warn) so it renders as a supported variant.
• Minor review-pass cleanup; no changes to guidance, formulas, or source baselines.

Affected Chapters

• versioning-release.mdx

2026-05-18

Change Summary

• Created the English edition from the Korean llmops-agentops handbook.
• Localized chapter titles, navigation links, formulas, callouts, evidence tables, and verification records.
• Preserved source baselines and freshness markers from the 2026-05-17 Korean 4th verification.

Affected Chapters

• index.mdx
• system-architecture.mdx
• versioning-release.mdx
• evaluation.mdx
• online-guardrails.mdx
• observability-slo.mdx
• cost-latency.mdx
• experimentation.mdx
• incident-management.mdx
• verification.mdx
• verification-archive.mdx
• updates.mdx

2026-05-17

Source Edition Change Summary

• Corrected A2A from v0.3.0 Draft to latest v1.0.0 (Ch1)
• Added MCP 2025-11-25 security requirements: OAuth 2.1, Protected Resource Metadata, Client ID Metadata Documents, token audience binding, token passthrough prohibition (Ch1)
• Reflected OpenAI Agents SDK operating surfaces: guardrails, human review, resumable approval state, hosted/private MCP, tracing, agent evals, voice agents (Ch3-Ch5)
• Added supply-chain, permission, and telemetry controls from OWASP MCP Top 10 and Agentic Skills Top 10 (Ch4, Ch8)
• Corrected OTel GenAI to Development and OWASP AOS to work-in-progress (Ch5)
• Refreshed model pricing around GPT-5.5/GPT-5.4/GPT-5.4 mini, Claude Opus/Sonnet/Haiku 4.x, and DeepSeek V4 Flash/Pro (Ch6)
• Fixed the λ_risk mismatch in the experiment decision formula (Ch7)
• Added MCP/Skill compromise, A2A abuse, and voice/realtime degradation incident types (Ch8)
• Added chapter-level source baselines, check:freshness, operational examples, and verification archive separation

Primary Sources

• https://developers.openai.com/api/docs/guides/agents
• https://developers.openai.com/api/docs/guides/agents/guardrails-approvals
• https://developers.openai.com/api/docs/guides/agents/integrations-observability
• https://developers.openai.com/api/docs/guides/agent-evals
• https://openai.com/api/pricing/
• https://modelcontextprotocol.io/specification/2025-11-25
• https://a2a-protocol.org/latest/specification/
• https://opentelemetry.io/docs/specs/semconv/gen-ai/
• https://aos.owasp.org/aos/
• https://owasp.org/www-project-mcp-top-10/
• https://owasp.org/www-project-agentic-skills-top-10/
• https://docs.claude.com/en/docs/test-and-evaluate/strengthen-guardrails/handle-streaming-refusals
• https://platform.claude.com/docs/en/about-claude/pricing
• https://api-docs.deepseek.com/quick_start/pricing/
• https://www.langchain.com/blog/introducing-langsmith-fleet
• https://www.braintrust.dev/docs/loop
• https://www.pagerduty.com/newsroom/pagerduty-expands-ai-ecosystem-to-supercharge-ai-agents/

Template

### YYYY-MM-DD

**Change Summary**

- ...

**Affected Chapters**

- `...`

**Primary Sources**

- ...