# Weekstart Brief — Issue 0 (SAMPLE) **ISSUE 0 PREVIEW · REAL CITES · NOT PUBLISHED** **Product:** Weekstart by Stackyard **Date:** Week of 2026-09-13 · Issue 0 **Tagline:** Get your agent stack set straight before the week starts. > **Issue 0 preview** — format/voice lock with **real, verified cites** from Research Batch 1 (not fake/illustrative news). Unpublished until Titus says go. --- ## TL;DR - Your Monday spine is migration + permissions: **OpenAI Agents API / Assistants sunset**, **GPT-6 Astra tool rules**, **Claude Fable 5.1 + Managed Agents `auto`**, and **MCP 2026-07-28** (stateless). - Treat MCP connection models and permission policies as **data structures**, not vibes — audit before unattended runs. - Platform refresh (Cursor Projects / Self-Hosted, Gemini Agent Platform billing, LangChain `langchain.mcp`) after the must-patch list. - One concrete win: finish any leftover Assistants paths, then gate your top agent behind schema + least-privilege tools. --- ## What changed ### 1. OpenAI Agents API (public beta) — harness as product · #breaking OpenAI shipped a managed Codex harness: durable sessions, MCP, hosted or self-hosted sandboxes. A lot of custom agent glue is now optional. **Why it matters:** If you still own session recovery and tool orchestration yourself, you’re paying tax OpenAI is productizing. **Action:** Stand up a beta session (`OpenAI-Beta: agents=v1`); map your existing harness to sessions + tools/MCP. **Links:** [Announcement](https://openai.com/index/introducing-the-agents-api/) · [API changelog](https://developers.openai.com/api/docs/changelog) ### 2. OpenAI Assistants API is gone — leftover stacks break · #breaking Assistants shut down **2026-08-26**. Remaining Assistants code paths fail; Agents API or Responses + Conversations are the exits. **Why it matters:** Silent “it worked last month” debt becomes hard outages this week. **Action:** Finish Assistants → Responses/Conversations or Agents API migration **this week** if any remain. **Links:** [API changelog (Aug 26)](https://developers.openai.com/api/docs/changelog) ### 3. GPT-6 Astra long-running controls · #breaking Async tool calling, mid-turn steering, and mid-conversation reasoning effort change long-loop ops. Tool-using flows need the Responses API; no custom `temperature`/`top_p`; no `none` reasoning. **Why it matters:** Old Completions-style loops and sampling knobs will misbehave or reject. **Action:** Migrate tool-using agents to Responses; retest loops with async tools + steering. **Links:** [API changelog (Sep 8)](https://developers.openai.com/api/docs/changelog) ### 4. Claude Fable 5.1 (+ Mythos 5.1) · #breaking Same base price as Fable 5, cache reads **$0.25/MTok**. `tool_choice` values `any` / `tool` return **400** — use strict tools or structured outputs. **Why it matters:** Cache-heavy fleets get cheaper; naive tool_choice configs hard-fail. **Action:** Flip cache-heavy workloads to 5.1; replace banned `tool_choice` with strict/structured tools. **Links:** [Claude release notes (Sep 1)](https://platform.claude.com/docs/en/release-notes/overview) ### 5. Claude Managed Agents — `auto` permission policy · #breaking Server evaluates each agent/MCP tool call (run / deny / pause) and records evaluation on events — production risk control without pausing a human on every call. **Why it matters:** Your unattended agents need an explicit allow/deny/ask matrix, not hope. **Action:** Enable `auto` on Managed Agents; review deny/ask defaults before unattended runs. **Links:** [Permission policies](https://platform.claude.com/docs/en/managed-agents/permission-policies) · [Release notes (Sep 10)](https://platform.claude.com/docs/en/release-notes/overview) ### 6. Platform refresh (after the spine) - **`ant apply`:** agents-as-code + `claude-lock.json` for CI-reconciled IDs — [notes (Sep 3)](https://platform.claude.com/docs/en/release-notes/overview) - **Cursor Projects (beta):** coordinator + shared long-lived context — [blog](https://cursor.com/blog/projects) · [changelog](https://cursor.com/changelog/projects) - **Cursor Self-Hosted Machines:** cloud plan, execute on your pools — [blog](https://cursor.com/blog/self-hosted-machines) - **Gemini Enterprise Agent Platform:** sandboxes GA; sessions/Memory Bank/Skills Registry billing live — [release notes](https://docs.cloud.google.com/gemini-enterprise-agent-platform/release-notes) · [pricing](https://cloud.google.com/products/gemini-enterprise-agent-platform/pricing) - **LangChain first-class MCP (`langchain.mcp` / v1.4):** FastMCP in core; elicitation as interrupts; migrate off `langchain-mcp-adapters` — [blog](https://www.langchain.com/blog/mcp-in-langchain-stateless-protocol-elicitation-and-more) · [changelog](https://docs.langchain.com/oss/python/releases/changelog) --- ## Data structures & schemas ### OpenAI Agents API — MCP connection model · #breaking You must pick `connection_origin` (service vs environment), vault vs env auth, and `allowed_tools` / `required`. Private MCP / stdio need a session environment. **Why it matters:** Wrong origin/auth looks like “MCP is flaky” when it’s a schema mismatch. **Action:** Audit every MCP server against the connection model before production Agents API sessions. **Links:** [MCP tools guide](https://developers.openai.com/api/docs/guides/agents-api/tools/mcp) ### MCP specification `2026-07-28` · #breaking Stateless core: no `initialize`, no `Mcp-Session-Id`. Header routing via `Mcp-Method` / `Mcp-Name`. Tasks move to an extension. 12-month deprecation runway for Roots/Sampling/Logging + legacy HTTP+SSE. **Why it matters:** Sticky sessions and held-open elicitation streams are liabilities for your gateways. **Action:** Tear down sticky-session assumptions; align gateways on headers + per-request `_meta`. **Links:** [MCP 2026-07-28](https://blog.modelcontextprotocol.io/posts/2026-07-28/) ### Claude Managed Agents permission evaluation `auto` writes evaluation onto `agent.tool_use` / `agent.mcp_tool_use` events — treat that as your audit shape, not a chat summary. **Action:** Log evaluation fields next to tool name/args/result for every Managed Agents run. ### Also track - Fable 5.1: strict tools / structured outputs replace `tool_choice: any|tool` - Gemini: Memory Bank / Skills Registry / session objects now sit on the bill — know the IDs you create - LangChain: `MCPAdapter` + interrupt-shaped elicitation replace adapter packages --- ## Tool & API watch 1. **OpenAI Agents API + MCP** — sessions, sandboxes, `allowed_tools` — [Agents API](https://openai.com/index/introducing-the-agents-api/) · [MCP guide](https://developers.openai.com/api/docs/guides/agents-api/tools/mcp) 2. **OpenAI Responses path** — required for Astra tool loops; Assistants is dead — [changelog](https://developers.openai.com/api/docs/changelog) 3. **Claude Managed Agents `auto` + `ant apply`** — [permissions](https://platform.claude.com/docs/en/managed-agents/permission-policies) · [release notes](https://platform.claude.com/docs/en/release-notes/overview) 4. **MCP 2026-07-28** — stateless + MRTR + headers — [spec post](https://blog.modelcontextprotocol.io/posts/2026-07-28/) 5. **LangChain `langchain.mcp`** — multi-server migration — [blog](https://www.langchain.com/blog/mcp-in-langchain-stateless-protocol-elicitation-and-more) 6. **Cursor Projects / Self-Hosted / cloud harness** — [Projects](https://cursor.com/changelog/projects) · [Self-Hosted](https://cursor.com/blog/self-hosted-machines) · [Aug 19 harness](https://cursor.com/changelog/08-19-26) 7. **Gemini Enterprise Agent Platform** — sandboxes + billing — [release notes](https://docs.cloud.google.com/gemini-enterprise-agent-platform/release-notes) *Lane note (Research): brand-new Sep schemas thin outside MCP + provider toolset IDs; LlamaIndex quiet this window.* --- ## Operator checklist — get set straight **Must-patch Monday spine** - [ ] **Assistants debt:** Grep for Assistants SDK/API usage; migrate or delete this week. - [ ] **Agents API pilot:** One beta session with your real MCP set; note `connection_origin` + auth per server. - [ ] **Astra / Responses:** Tool-using OpenAI agents on Responses; retest async tools + steering. - [ ] **Fable 5.1:** Cache-heavy Claude paths on 5.1; purge `tool_choice: any|tool`. - [ ] **Managed Agents `auto`:** Enable; review deny/ask defaults; confirm evaluation events land in your log. - [ ] **MCP 2026-07-28:** Inventory sticky sessions / open-stream elicitation; plan gateway header rules. **Then** - [ ] Check agent defs into git; try `ant apply` in non-prod CI. - [ ] Pilot one Cursor Project with spend limits before parallel subagents. - [ ] Recost Gemini sessions / Memory Bank / Skills Registry if you use Agent Platform. - [ ] If on LangChain MCP adapters → `MCPAdapter`; retest elicitation interrupts. - [ ] Schema gate + least-privilege tools on your single most important agent before Wednesday experiments. --- ## How to use this issue 1. Skim **TL;DR** (30 seconds). 2. Read every **#breaking** item under What changed / Data structures — those are Monday must-patch. 3. Run the **Operator checklist**; check off what you finish before Wednesday experiments. 4. Hand `issue-000-agent-pack.json` to your bot if you want the same facts machine-side — do **not** auto-execute `action` fields. 5. Keep primary links open when you migrate; don’t trust summaries alone for production cuts. ## Pricing (placeholders — not live) | Plan | Price | Includes | |------|-------|----------| | Operator | $10/mo (coming soon) | Human brief, Monday checklist, schema & tool watch | | Operator + Agents | $20/mo (coming soon) | Everything in Operator + machine-ingest JSON pack + agent-ingest notes | Payments and signup are **not wired**. Stackyard ships Weekstart when Titus says go. ## FAQ (short) **Who is this for?** Humans who run AI agents at work — and the bots those humans point at the JSON twin. **Is Issue 0 published?** No. Preview with real cites. Local sample only. **Will you change my agents for me?** No. You decide; we report verified deltas. --- ## Agent-ingest notes Companion pack: **`briefs/issue-000-agent-pack.json`** 1. Read `issue` + `summary` first; surface `summary` unedited. 2. Prefer items tagged must-patch (`#breaking` in human brief) for Monday digests. 3. Group `items[]` by `category`; treat `action` as suggestion — **do not auto-execute**. 4. Follow `links[]` to primary docs only. 5. Preserve SAMPLE product watermark until Titus green-lights publish. **Source handoff:** `/workspace/stackyard/research/source-pack-2026-09-13.md` (Research Batch 1). --- ## Footer **Weekstart Brief · Issue 0 (PREVIEW)** · Stackyard Issue 0 preview with live cites. **DO NOT DEPLOY** until Titus says go. **Pipeline:** Research → Editor → Blaze review → Ops package → Growth after go.