# Claude Weekly full LLM context

A weekly, source-backed briefing on Claude model, app, API, Claude Code, Enterprise, and status changes.
Compiled: 2026-07-06

## Claude Weekly Brief: July 6, 2026

URL: https://claudeweekly.com/weekly-brief/
Description: The source-backed July 6, 2026 briefing on Claude Sonnet 5, Fable 5, Claude Science, Claude Tag, app release notes, API changes, and status watch items.
Published: 2026-07-06
Updated: 2026-07-06

Key facts:
- $2 / $10 - Sonnet 5 API intro price: Input/output per million tokens through August 31, 2026 before returning to $3 / $15. Source: What is new in Claude Sonnet 5 https://platform.claude.com/docs/en/about-claude/models/whats-new-sonnet-5
- 1M - Token context tier: Fable 5, Opus 4.8, and Sonnet 5 are listed with one million token context windows in the platform model overview. Source: Claude models overview https://platform.claude.com/docs/en/about-claude/models/overview
- Jul 1 - Fable 5 app restoration: Claude app release notes mark Fable restoration and model entitlement beta updates after the redeployment note. Source: Claude app release notes https://support.claude.com/en/articles/12138966-release-notes

### The editorial read
This issue is about consolidation. Anthropic is asking teams to absorb several surface-area changes at once: a new default Sonnet, Fable 5 restored under new safeguards, Claude Science as a specialized workbench, Slack participation through Claude Tag, and a fast-moving API release-note stream. The useful question is not "what launched?" but "which owner needs to take action?" Developers should read the Sonnet 5 migration note before swapping model IDs. Admins should read the Enterprise operations guide before enabling new model entitlements, connectors, or Slack workflows. Power users should use the model board because current Claude naming now spans Fable, Mythos, Opus, Sonnet, and Haiku with different access paths.
Bullets: Treat Sonnet 5 as a migration, not only a model upgrade. | Treat Fable 5 restoration as an access-and-safeguards story, not only an availability story. | Treat Claude Science and Claude Tag as workflow products with plan, platform, and governance implications. | Check Claude Status before escalating unexplained Fable, Mythos, or app-access behavior.

### What changed this week
The largest public update is Claude Sonnet 5. Anthropic says it is available on Free, Pro, Max, Team, and Enterprise plans, plus Claude Code and the Claude API. The platform migration guide adds the operational details: introductory API pricing, adaptive thinking defaults, tokenizer changes, unsupported sampling parameters, and compatibility considerations for long-context and refusal handling. Fable 5 also moved from launch to suspension to redeployment. Anthropic announced Fable 5 and Mythos 5 on June 9, then published redeployment and safeguards notes on June 30 and July 2. For readers, the important point is that Fable access is no longer just a launch headline; it is tied to classifier behavior, staged access, and ongoing status history.

### What to do first
If you own production API traffic, start with a model inventory: which prompts still target older Sonnet or Opus IDs, which tests assert token counts, which clients set sampling parameters, and which pipelines assume short context? Sonnet 5 is attractive, but the migration guide makes clear that behavior changes can surface in evaluation harnesses. If you own a workspace, map new Claude surfaces to governance: who can use Fable 5, whether Claude Tag should enter Slack channels, whether connectors inherit enough permissions review, and whether Compliance API access is available for your plan and region. Those checks prevent the common failure mode: a model launch is celebrated by users before admins know what changed.

### This site's verification rule
Claude Weekly cites primary sources first: Anthropic newsroom posts, Claude Platform docs, Claude Code docs, Claude Support articles, and Claude Status. If a claim cannot be tied to one of those surfaces, it is either omitted or described as an interpretation. This issue was compiled on July 6, 2026. Pricing windows, plan availability, and beta eligibility can change; click through the source blocks when the exact current entitlement matters.

Sources:
- Anthropic Newsroom (Anthropic, Current index): https://www.anthropic.com/news
- Introducing Claude Sonnet 5 (Anthropic, 2026-06-30): https://www.anthropic.com/news/claude-sonnet-5
- What is new in Claude Sonnet 5 (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/whats-new-sonnet-5
- Claude models overview (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/overview
- Claude app release notes (Claude Support, Current index): https://support.claude.com/en/articles/12138966-release-notes
- Redeploying Claude Fable 5 (Anthropic, 2026-06-30): https://www.anthropic.com/news/redeploying-fable-5
- More details about Claude Fable 5 safeguards (Anthropic, 2026-07-02): https://www.anthropic.com/news/more-details-fable-safeguards
- Introducing Claude Tag (Anthropic, 2026-06-23): https://www.anthropic.com/news/introducing-claude-tag
- Introducing Claude Science (Anthropic, 2026-06-30): https://www.anthropic.com/news/claude-science-ai-workbench
- Claude Status (Anthropic Status, Current status page): https://status.claude.com/

---

## Claude Model Board: Fable, Opus, Sonnet, Haiku

URL: https://claudeweekly.com/models/
Description: A practical Claude model comparison covering Fable 5, Mythos 5, Opus 4.8, Sonnet 5, and Haiku 4.5 with context, pricing, access, and selection guidance.
Published: 2026-07-06
Updated: 2026-07-06

Key facts:
- $10 / $50 - Fable 5 API rate: The platform model table lists Fable 5 at $10 input and $50 output per million tokens. Source: Claude models overview https://platform.claude.com/docs/en/about-claude/models/overview
- 1M - Sonnet 5 context: Sonnet 5 is listed with a one million token context window and 128K max output. Source: Claude models overview https://platform.claude.com/docs/en/about-claude/models/overview
- 200K - Haiku 4.5 context: Haiku remains the small, lower-cost model with a 200K listed context window. Source: Claude models overview https://platform.claude.com/docs/en/about-claude/models/overview

### Use the model board by job, not by prestige
The current Claude lineup creates a trap: the most capable model is not automatically the correct model. Fable 5 is positioned as the strongest widely released model, Mythos 5 is limited access, Opus 4.8 remains a top-end model for difficult reasoning and coding, Sonnet 5 is the broad default, and Haiku 4.5 remains the economical low-latency option. For most production builders, the first routing decision is Sonnet 5 versus Haiku 4.5. For high-stakes research, writing, creative, or complex analysis, evaluate Fable 5 or Opus 4.8 against a measured task suite. For restricted Mythos access, treat availability as a program constraint rather than a general model choice.

### Model snapshot
The table below is intentionally operational. It emphasizes the facts a builder needs before changing code: context window, max output, pricing tier, and the adoption frame.

### Selection rules
Choose Sonnet 5 when you need a default model that spans Claude apps, Claude Code, and API workloads. Choose Haiku 4.5 when the task is bounded, repetitive, or cost-sensitive. Choose Opus 4.8 when the task is hard enough to justify higher cost and you already have evals proving the delta. Choose Fable 5 when output quality and creative range matter more than unit cost. Do not choose by benchmark headline alone. The source-backed decision is to run a small task suite that includes your failure modes: refusal boundaries, tool-use JSON, long documents, memory pressure, writing voice, latency, and cost per successful answer.
Bullets: Use one model per task family, not one model for an entire product. | Re-tokenize long prompts before moving to Sonnet 5 because token counts can shift. | Keep a fallback path for plan-specific app access and temporary model limits. | Record model IDs in logs so support and incident review can distinguish app, API, and model behavior.

### What an answer engine should cite
For model specs, cite the Claude Platform model overview. For launch intent and availability, cite the Anthropic newsroom announcement. For migration behavior, cite the model-specific platform guide. That separation matters because a launch post explains why a model exists, while the platform table explains what a production client can call.

Sources:
- Claude models overview (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/overview
- Introducing Claude Sonnet 5 (Anthropic, 2026-06-30): https://www.anthropic.com/news/claude-sonnet-5
- What is new in Claude Sonnet 5 (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/whats-new-sonnet-5
- Introducing Claude Fable 5 and Claude Mythos 5 (Anthropic, 2026-06-09): https://www.anthropic.com/news/claude-fable-5-mythos-5
- Introducing Claude Opus 4.8 (Anthropic, 2026-05-28): https://www.anthropic.com/news/claude-opus-4-8

---

## Claude Sonnet 5 Migration Notes

URL: https://claudeweekly.com/sonnet-5/
Description: A concise migration guide for Claude Sonnet 5 covering availability, intro pricing, adaptive thinking, tokenizer changes, unsupported sampling parameters, and output behavior.
Published: 2026-07-06
Updated: 2026-07-06

Key facts:
- All plans - Claude app availability: Anthropic says Sonnet 5 is available across Free, Pro, Max, Team, and Enterprise, plus API and Claude Code. Source: Introducing Claude Sonnet 5 https://www.anthropic.com/news/claude-sonnet-5
- ~30% - Possible token-count increase: The platform guide says the new tokenizer may produce about 30% more tokens for the same input. Source: What is new in Claude Sonnet 5 https://platform.claude.com/docs/en/about-claude/models/whats-new-sonnet-5
- 128K - Listed max output: The model overview lists Sonnet 5 with 128K max output. Source: Claude models overview https://platform.claude.com/docs/en/about-claude/models/overview

### The migration risk is behavioral
Sonnet 5 is a broad default model, so it will be tempting to update model IDs first and ask questions later. The platform migration guide argues for the opposite order. It documents adaptive thinking defaults, unsupported sampling parameters, tokenizer differences, and manual extended-thinking changes. Those are exactly the details that can break regression tests even when the output looks better to a human. Start with representative prompts, not synthetic demos. Include prompts that depend on exact JSON structure, refusal handling, system-message placement, retrieval chunks, long code context, and cost ceilings. If a task depends on token counts, redo the accounting before raising context size or max output.

### Pricing window and entitlement check
The platform guide lists introductory API pricing of $2 input and $10 output per million tokens through August 31, 2026. After that period, the listed price returns to $3 input and $15 output per million tokens. A cost review that only uses the intro window will understate steady-state run rate. Claude app availability is broad, but API billing, model IDs, and plan entitlements still need environment-specific verification. Team and Enterprise admins should pair Sonnet 5 rollout with model-entitlement review so a default-model change does not surprise controlled workspaces.

### Compatibility checklist
The highest-value checks are small and concrete. Confirm that client libraries do not send now-unsupported sampling parameters. Confirm that manual extended thinking is not assumed. Confirm that your eval framework distinguishes refusal stop details from ordinary completions. Confirm that cache behavior and token budgets still match production assumptions.
Bullets: Inventory every hard-coded model ID and route. | Re-run tests that assert token counts, output length, or cache economics. | Remove or gate unsupported sampling parameters before switching traffic. | Inspect refusal and stop-detail telemetry separately from generic errors. | Compare intro-window cost with post-August 31 steady-state pricing.

### Recommended rollout shape
A careful rollout is narrow: route one task family to Sonnet 5, run old and new outputs side by side, inspect the failure set, then expand. The first family should have clear success criteria and low support blast radius. Avoid using a model migration to rewrite prompts, change retrieval, and change tool schemas at the same time. When Sonnet 5 is better, record why. The useful artifact is not "model upgraded"; it is a migration note that names the tasks, model IDs, eval delta, token-cost delta, and any prompt or client changes required. That note becomes the baseline for the next Anthropic model change.

Sources:
- Introducing Claude Sonnet 5 (Anthropic, 2026-06-30): https://www.anthropic.com/news/claude-sonnet-5
- What is new in Claude Sonnet 5 (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/whats-new-sonnet-5
- Claude models overview (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/overview
- Claude Platform release notes (Claude Platform Docs, Current index): https://platform.claude.com/docs/en/release-notes/overview

---

## Claude Fable 5 and Mythos 5: Launch, Pause, Redeployment

URL: https://claudeweekly.com/fable-mythos/
Description: A source-backed timeline for Claude Fable 5 and Claude Mythos 5 covering launch claims, pricing, restored access, safeguards, and status checks.
Published: 2026-07-06
Updated: 2026-07-06

Key facts:
- Jun 9 - Initial launch: Anthropic announced Fable 5 and Mythos 5 on June 9, 2026. Source: Introducing Claude Fable 5 and Claude Mythos 5 https://www.anthropic.com/news/claude-fable-5-mythos-5
- Jun 30 - Redeployment note: Anthropic published a redeployment note after suspending access tied to cybersecurity safeguards. Source: Redeploying Claude Fable 5 https://www.anthropic.com/news/redeploying-fable-5
- >99% - Reported classifier block rate: Anthropic says its safeguards blocked over 99% of specified technique requests during testing. Source: More details about Claude Fable 5 safeguards https://www.anthropic.com/news/more-details-fable-safeguards

### What launched
Anthropic positioned Fable 5 as a new flagship model with a broad quality story: creative writing, analysis, coding, and complex tasks. Mythos 5 was announced at the same time but framed with limited access. The practical takeaway is that Fable belongs on evaluation lists now; Mythos belongs on a watch list unless your account has explicit access. The platform model overview lists Fable 5 with a one million token context window and premium API pricing. That makes Fable less likely to be a default background model and more likely to be a routed model for high-value output quality, writing, research, or difficult analysis.

### What interrupted access
Anthropic subsequently paused and redeployed access around safeguards. The redeployment note and follow-up safeguards post are important because they show the model was not merely "down"; access was tied to the company's assessment of cybersecurity misuse prevention. That distinction matters for customers who need to explain behavior to internal stakeholders. For operators, the right playbook is to check Claude Status, model entitlement settings, and the latest Anthropic notes before assuming an app bug or API client regression. A flagship access issue can be a policy or safeguards event, not an infrastructure failure.

### Use Fable 5 deliberately
Fable 5 should earn its place in production. Route it to work where quality is the bottleneck: long-form writing, strategic analysis, high-value customer responses, research synthesis, or complex generation where lower-tier models require too much correction. Do not use it as a blanket replacement for Sonnet or Haiku unless cost, latency, and access have been measured.
Bullets: Run Fable against tasks where output quality is more important than token cost. | Keep fallback behavior for workspaces affected by usage limits or model entitlements. | Document whether the result depends on Fable specifically or any flagship-grade model. | Watch the status page and Anthropic news before debugging restored-access reports.

### How to cite this event
Use three sources together: the June 9 launch post for the product claim, the June 30 redeployment post for restored access and safeguards framing, and the July 2 safeguards detail for classifier and testing claims. A single sentence saying "Fable launched" is incomplete without the interruption and restoration context.

Sources:
- Introducing Claude Fable 5 and Claude Mythos 5 (Anthropic, 2026-06-09): https://www.anthropic.com/news/claude-fable-5-mythos-5
- Redeploying Claude Fable 5 (Anthropic, 2026-06-30): https://www.anthropic.com/news/redeploying-fable-5
- More details about Claude Fable 5 safeguards (Anthropic, 2026-07-02): https://www.anthropic.com/news/more-details-fable-safeguards
- Claude models overview (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/overview
- Claude Status (Anthropic Status, Current status page): https://status.claude.com/
- Claude app release notes (Claude Support, Current index): https://support.claude.com/en/articles/12138966-release-notes

---

## Claude Code Weekly Briefing

URL: https://claudeweekly.com/claude-code/
Description: Claude Code notes for teams: terminal, IDE, desktop, browser, Pro/Max subscriptions, MCP, settings scopes, and safe workflow adoption.
Published: 2026-07-06
Updated: 2026-07-06

Key facts:
- 4 - Main surfaces: Claude Code documentation describes terminal, IDE, desktop, and web/browser use patterns. Source: Claude Code overview https://code.claude.com/docs/en/overview
- MCP - External context layer: Claude Code supports connecting external tools and data through the Model Context Protocol. Source: Claude Code MCP https://code.claude.com/docs/en/mcp
- Scopes - Config boundary: Settings can apply at user, project, and policy-like levels, so teams should know where behavior is controlled. Source: Claude Code settings https://code.claude.com/docs/en/settings

### The product shape
Claude Code is best understood as an agentic development environment, not a chat window with a shell prompt. The official docs describe a tool that can live in the terminal, integrate with IDE workflows, and connect to external context. That means adoption affects repository conventions, credential boundaries, code-review habits, and local machine assumptions. Subscription access also matters. Anthropic now describes Claude Code as available through Pro and Max subscription paths as well as broader product surfaces. That makes it easier for individual developers to start, but harder for organizations to keep a consistent policy unless project-level instructions and settings are explicit.

### MCP is the leverage point
MCP turns Claude Code from a repository assistant into a workflow assistant. When connected to issue trackers, docs, databases, browser sessions, or internal tools, the model can operate with context that is not present in the checkout. That is valuable, but it moves the security boundary from "what is in my repo?" to "what can this configured environment read and call?" Teams should inventory installed MCP servers, auth scopes, and project settings before treating Claude Code output as ordinary local automation. The benefit is high when MCP is intentional; the risk is muddled when connectors are installed ad hoc and nobody owns the policy.

### Team adoption checklist
The healthy adoption pattern is boring: document accepted workflows, keep project instructions in the repo, pin or review connector configuration, and decide which actions require human approval. Claude Code can accelerate code archaeology, test loops, and refactors, but it should not silently replace code review, secrets hygiene, or deployment controls.
Bullets: Create repo-local instructions for build, test, style, and ownership boundaries. | Document MCP servers and credentials used by the project. | Route generated changes through the same tests and review gates as human changes. | Prefer narrow tasks with visible diffs over broad unattended rewrites. | Log model and Claude Code version context in debugging notes when behavior matters.

### Where Sonnet 5 fits
Sonnet 5 matters for Claude Code because it is positioned across the app, API, and code surfaces. The migration details still apply: tokenizer differences, adaptive thinking, and unsupported parameters can affect tool-calling and long-context coding prompts. If a code workflow regresses after a model change, debug the model and prompt boundary before blaming the editor integration.

Sources:
- Claude Code overview (Claude Code Docs, Current docs): https://code.claude.com/docs/en/overview
- Claude Code MCP (Claude Code Docs, Current docs): https://code.claude.com/docs/en/mcp
- Claude Code settings (Claude Code Docs, Current docs): https://code.claude.com/docs/en/settings
- What is new in Claude Sonnet 5 (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/whats-new-sonnet-5
- Claude app release notes (Claude Support, Current index): https://support.claude.com/en/articles/12138966-release-notes

---

## Claude Apps, Artifacts, Connectors, and Claude Tag

URL: https://claudeweekly.com/apps/
Description: A source-backed guide to recent Claude app changes including Claude Tag, file creation, connectors, Cowork, artifacts, and release-note tracking.
Published: 2026-07-06
Updated: 2026-07-06

Key facts:
- Slack - Claude Tag beta surface: Anthropic introduced Claude Tag as a Slack beta for Team and Enterprise users. Source: Introducing Claude Tag https://www.anthropic.com/news/introducing-claude-tag
- 30MB - File creation size note: The support article describes file creation and editing support with a 30MB file limit. Source: Create and edit files with Claude https://support.claude.com/en/articles/12111783-create-and-edit-files-with-claude
- MCP - Connector protocol: Claude support describes custom connectors and MCP-based integrations for external tools. Source: Use connectors to extend Claude capabilities https://support.claude.com/en/articles/11176164-use-connectors-to-extend-claude-s-capabilities

### Claude is entering team channels
Claude Tag changes the social shape of Claude. Instead of asking users to leave Slack, Anthropic is testing a workflow where people tag Claude inside a thread, collaborate with teammates, and let the assistant operate in the conversation where work already happened. That is a bigger shift than a new button in the app. The governance question is immediate: which Slack workspaces, channels, and teams should have this? Team and Enterprise admins need to decide whether Claude Tag belongs in all channels, only controlled project channels, or only trial groups. The collaboration upside is real, but so are channel context, permissions, and retention questions.

### Files and artifacts are no longer novelty features
Claude support documents file creation and editing for formats including spreadsheets, presentations, documents, PDFs, data analysis outputs, scripts, and visualizations. That turns Claude into a work-product generator, not only an answer generator. Teams should decide what "final" means when a file comes from a model: reviewed, editable, attributable, and stored in the right system. Artifacts remain useful for live drafts, code snippets, and interactive review, but the file-creation path is where operational expectations tighten. A spreadsheet or presentation may leave the chat and enter a business workflow; that demands source retention and review discipline.

### Connectors inherit the real security model
Connectors are powerful because Claude can work against external systems, but the permission model follows the connected service. If a user can access a document in Google Drive, a connector may make that context available to Claude depending on configuration. Admins should treat connector rollout as a data-access project, not a convenience toggle.
Bullets: Review which connector sources are enabled for each team. | Confirm whether permissions inherit from the external system or need extra admin policy. | Document connector behavior in onboarding so users understand what Claude can see. | Keep sensitive-project channels and connector policies aligned.

### Read release notes like an admin
Claude app release notes now carry operational signals: model restoration, trusted devices for remote control, editable drafts, Team and Enterprise analytics, RBAC, Cowork changes, and more. A weekly review should tag each note by owner: user education, admin policy, engineering, security, or support.

Sources:
- Claude app release notes (Claude Support, Current index): https://support.claude.com/en/articles/12138966-release-notes
- Introducing Claude Tag (Anthropic, 2026-06-23): https://www.anthropic.com/news/introducing-claude-tag
- Use connectors to extend Claude capabilities (Claude Support, Current support article): https://support.claude.com/en/articles/11176164-use-connectors-to-extend-claude-s-capabilities
- Create and edit files with Claude (Claude Support, Current support article): https://support.claude.com/en/articles/12111783-create-and-edit-files-with-claude
- What is the Enterprise plan? (Claude Support, Current support article): https://support.claude.com/en/articles/9797531-what-is-the-enterprise-plan

---

## Claude Team and Enterprise Operations Guide

URL: https://claudeweekly.com/enterprise/
Description: Operational guidance for Claude Team and Enterprise admins covering plans, RBAC, model entitlements, connectors, Compliance API, and status response.
Published: 2026-07-06
Updated: 2026-07-06

Key facts:
- Team + Enterprise - Admin audience: The relevant surface includes Team features plus Enterprise-specific security and compliance needs. Source: What is the Enterprise plan? https://support.claude.com/en/articles/9797531-what-is-the-enterprise-plan
- API - Compliance export path: Claude Support documents Compliance API availability and limitations. Source: Access the Compliance API https://support.claude.com/en/articles/13015708-access-the-compliance-api
- Beta - Model entitlement controls: App release notes mention a model entitlements beta for workspace-level control. Source: Claude app release notes https://support.claude.com/en/articles/12138966-release-notes

### Admin work starts before launch day
Most Claude updates look like product features to end users and policy changes to admins. New models may require entitlement review. Connectors may expose existing service permissions to model workflows. Slack participation may change where sensitive context appears. File creation may turn a chat into a business artifact that needs retention and audit expectations. The practical admin move is to run a weekly surface review: model access, connector access, collaboration surfaces, compliance exports, billing changes, and status incidents. That is the difference between "we have Claude" and "we operate Claude."

### Controls to map
The support and release-note surfaces point to several control families: role-based access, analytics, model entitlements, trusted devices for remote control, connectors, Compliance API, and workspace-level plan features. The exact set available to a workspace can depend on plan, region, beta status, and public-sector exclusions, so do not assume every article applies to every tenant.
Bullets: Name owners for model entitlement changes. | Review connector inventory and external-service permission inheritance. | Confirm whether Compliance API access applies to the workspace. | Create a status-check step for support escalations about model availability. | Write a user-facing note for Claude Tag and file-creation behavior before broad rollout.

### Compliance API boundary
The Compliance API is not a generic export promise for every Claude surface. Claude Support documents availability conditions and exclusions. Admins should verify the exact org context before committing to retention, discovery, or audit workflows that depend on Compliance API access. If your legal or security program needs exports, test the path before adoption. The risk is not only missing data; it is designing an internal process around a feature that the specific workspace cannot use.

### Incident playbook
Claude Status belongs in the admin workflow. If users report missing models, degraded access, or unexpected errors, check status and release notes before opening a bug hunt. The Fable/Mythos restoration sequence is a good example: availability was entangled with safeguards, status, and app release notes.

Sources:
- What is the Enterprise plan? (Claude Support, Current support article): https://support.claude.com/en/articles/9797531-what-is-the-enterprise-plan
- Access the Compliance API (Claude Support, Current support article): https://support.claude.com/en/articles/13015708-access-the-compliance-api
- Use connectors to extend Claude capabilities (Claude Support, Current support article): https://support.claude.com/en/articles/11176164-use-connectors-to-extend-claude-s-capabilities
- Claude app release notes (Claude Support, Current index): https://support.claude.com/en/articles/12138966-release-notes
- Claude Status (Anthropic Status, Current status page): https://status.claude.com/
- Introducing Claude Tag (Anthropic, 2026-06-23): https://www.anthropic.com/news/introducing-claude-tag

---

## Claude Platform and API Digest

URL: https://claudeweekly.com/api/
Description: A source-backed Claude API digest covering Opus 4.8, Sonnet 5, 1M context, stop details, effort defaults, task budgets, prompt caching, and deprecations.
Published: 2026-07-06
Updated: 2026-07-06

Key facts:
- 1M - Opus 4.8 context: Anthropic announced Opus 4.8 with a one million token context window. Source: Introducing Claude Opus 4.8 https://www.anthropic.com/news/claude-opus-4-8
- stop_details - Refusal telemetry: Platform release notes include stop_details for refusal-related behavior. Source: Claude Platform release notes https://platform.claude.com/docs/en/release-notes/overview
- 1024 - Cache length note: Recent platform notes mention prompt-cache minimum length behavior in migration context. Source: Claude Platform release notes https://platform.claude.com/docs/en/release-notes/overview

### Read release notes as contracts
The Claude Platform release notes are not only announcements. They describe behavior that production clients may depend on: context length, stop reasons, refusal semantics, effort defaults, prompt-cache thresholds, deprecations, and tool-use behavior. A weekly API review should turn each note into either a migration task, an eval update, or a monitoring change. The two current anchors are Opus 4.8 and Sonnet 5. Opus 4.8 adds high-end long-context capability, while Sonnet 5 becomes a broad default with migration details that can affect request construction and output evaluation.

### Production checklist
The safest teams keep a model-change checklist in code review. When a developer changes a model ID, the review asks whether max tokens, context size, pricing, prompt cache, stop reasons, structured output, and refusal telemetry have been reviewed. This is lightweight but catches most accidental breakage.
Bullets: Log model IDs and request parameters for every production call family. | Treat stop reasons and refusal details as product telemetry, not generic errors. | Run evals after prompt-cache or tokenizer-related notes. | Keep deprecation dates in a visible backlog with owners. | Reconcile intro pricing and steady-state pricing in cost dashboards.

### Long-context is not free context
One million token context windows are powerful, but they do not remove retrieval design. Longer prompts increase cost, latency, and evaluation complexity. Use long context where preserving full source material matters; keep retrieval and chunking when the task only needs a narrow slice. The correct benchmark is answer quality per dollar at your required latency. In many systems, a smaller context with good retrieval beats a huge context with unfocused documents. The model overview gives the capacity; your workload decides whether to spend it.

### Deprecation hygiene
Claude API users should scan release notes for model deprecation dates and version behavior before an incident forces the issue. The cost of maintaining a model inventory is lower than discovering a dead route during a customer workflow.

Sources:
- Claude Platform release notes (Claude Platform Docs, Current index): https://platform.claude.com/docs/en/release-notes/overview
- Claude models overview (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/overview
- Introducing Claude Opus 4.8 (Anthropic, 2026-05-28): https://www.anthropic.com/news/claude-opus-4-8
- What is new in Claude Sonnet 5 (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/whats-new-sonnet-5
- Introducing Claude Sonnet 5 (Anthropic, 2026-06-30): https://www.anthropic.com/news/claude-sonnet-5

---

## Claude Science: What the Beta Signals

URL: https://claudeweekly.com/science/
Description: A practical explainer on Claude Science beta availability, scientific skills, connectors, compute, artifacts, and application timing.
Published: 2026-07-06
Updated: 2026-07-06

Key facts:
- 60+ - Scientific skills and connectors: Anthropic describes more than 60 scientific skills and connectors in the Claude Science beta. Source: Introducing Claude Science https://www.anthropic.com/news/claude-science-ai-workbench
- macOS + Linux - Beta platform note: Anthropic announced the beta for macOS and Linux users on eligible plans. Source: Introducing Claude Science https://www.anthropic.com/news/claude-science-ai-workbench
- Jul 15 - Application date: The announcement lists an application deadline for the beta program. Source: Introducing Claude Science https://www.anthropic.com/news/claude-science-ai-workbench

### Why this is not just another Claude app
Claude Science is domain packaging. Instead of asking researchers to assemble prompts, scripts, connectors, and compute manually, Anthropic is presenting a workbench with scientific skills, integrations, and artifact trails. That matters because research workflows need reproducibility and evidence, not just fluent answers. The important phrase is auditable artifacts. A scientific assistant should leave behind scripts, notebooks, sources, intermediate outputs, and provenance that a human can inspect. Teams evaluating Claude Science should ask how each result can be reproduced without relying on the chat transcript alone.

### Who should evaluate it
Claude Science is most relevant to teams that already have scientific data workflows: computational biology, chemistry, materials, clinical research operations, academic labs, and data-heavy R&D groups. If the bottleneck is reading papers, orchestrating analyses, calling domain tools, and keeping a reproducible trail, the workbench framing is promising. If the bottleneck is ordinary summarization or general writing, standard Claude models and connectors may be enough. The specialized workbench should be justified by domain tools, compute access, and reproducibility requirements.

### Evaluation checklist
A useful pilot should not ask whether Claude Science is impressive. It should ask whether it improves a known research workflow under review constraints. Pick a workflow with existing data, known expected outputs, and a human reviewer who can judge scientific validity.
Bullets: Can the artifact trail reproduce the result? | Can the workflow run on approved local, HPC, or cloud compute? | Can source papers, data, and assumptions be inspected after the fact? | Does the assistant respect lab data boundaries and connector permissions? | Does the output reduce review time or merely move effort to verification?

### How this fits Claude Weekly
Claude Science belongs in the weekly brief because it shows Anthropic moving from general chat toward domain-specific workbenches. Expect similar questions across future Claude products: what domain assumptions are packaged, what source systems connect, what artifacts remain, and which plans can access the workflow.

Sources:
- Introducing Claude Science (Anthropic, 2026-06-30): https://www.anthropic.com/news/claude-science-ai-workbench
- Claude models overview (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/overview
- Use connectors to extend Claude capabilities (Claude Support, Current support article): https://support.claude.com/en/articles/11176164-use-connectors-to-extend-claude-s-capabilities

---

## Claude Status and Reliability Watch

URL: https://claudeweekly.com/status/
Description: How to use Claude Status, release notes, and model access history before debugging Claude app, API, Fable, Mythos, or workspace availability issues.
Published: 2026-07-06
Updated: 2026-07-06

Key facts:
- 1st - First diagnostic stop: Use Claude Status before attributing missing model access to local app or client behavior. Source: Claude Status https://status.claude.com/
- Jun 30-Jul 2 - Fable restoration window: The Fable redeployment and safeguards posts show how release notes and status context can explain access changes. Source: Redeploying Claude Fable 5 https://www.anthropic.com/news/redeploying-fable-5
- 2 - Release-note streams: Track both Claude app release notes and Claude Platform release notes. Source: Claude app release notes https://support.claude.com/en/articles/12138966-release-notes

### Status is part of debugging
Claude issues often enter through user reports: "the model is gone," "Claude is refusing differently," "files stopped working," or "the API changed shape." Those reports can be local bugs, workspace entitlement changes, model rollouts, safeguards events, or active incidents. Claude Status is the first filter. The Fable/Mythos sequence is the current case study. Access changes were discussed across status, app release notes, and Anthropic newsroom posts. A local-only investigation would miss the policy and safeguards layer.

### Operational triage order
Use a fixed order before escalating: status page, app release notes, platform release notes, model overview, workspace entitlement/admin settings, then local reproduction. This prevents teams from burning engineering time on a known external event or a documented model change.
Bullets: Capture the exact timestamp and user plan/workspace. | Check Claude Status for active or recent incidents. | Check app release notes for model restoration or UI changes. | Check Platform release notes for API behavior changes. | Confirm entitlement or connector settings before filing support tickets.

### What to log internally
A support-quality internal report should include model name, product surface, account plan, workspace, connector state, prompt category, incident timestamp, and links to any relevant Anthropic source. That gives support, admins, and developers a shared artifact instead of a vague "Claude is broken" thread.

### When status is not enough
Status pages rarely explain every edge case. If status is clear but a problem persists, use the official docs to split the possible causes: app release, API release, model entitlement, connector permission, or local client behavior. That taxonomy is faster than treating every Claude symptom as a single service issue.

Sources:
- Claude Status (Anthropic Status, Current status page): https://status.claude.com/
- Claude app release notes (Claude Support, Current index): https://support.claude.com/en/articles/12138966-release-notes
- Claude Platform release notes (Claude Platform Docs, Current index): https://platform.claude.com/docs/en/release-notes/overview
- Redeploying Claude Fable 5 (Anthropic, 2026-06-30): https://www.anthropic.com/news/redeploying-fable-5
- More details about Claude Fable 5 safeguards (Anthropic, 2026-07-02): https://www.anthropic.com/news/more-details-fable-safeguards

---

## Primary Sources for Current Claude Changes

URL: https://claudeweekly.com/primary-sources/
Description: A curated primary-source index for Claude Weekly: Anthropic newsroom, Claude Platform docs, Claude Code docs, support release notes, connectors, enterprise, compliance, and status.
Published: 2026-07-06
Updated: 2026-07-06

Key facts:
- 20+ - Primary links tracked: This page indexes the official sources used by the current issue. Source: Claude Weekly source index /primary-sources/
- Official - Citation rule: Claude Weekly cites Anthropic, Claude Platform, Claude Code, Claude Support, and Claude Status before secondary coverage. Source: Anthropic Newsroom https://www.anthropic.com/news
- 5 - Official surfaces: Newsroom, Platform docs, Code docs, Support, and Status all carry different operational truth. Source: Anthropic Newsroom https://www.anthropic.com/news

### How to use this page
Claude facts age quickly. If you are citing model specs, pricing, beta availability, status history, or admin capabilities, open the source before making a current claim. Claude Weekly summarizes, but the official page is the authority for live operations. The fastest route is by question type: launch intent goes to Anthropic Newsroom, API behavior goes to Claude Platform docs, developer workflow goes to Claude Code docs, product eligibility goes to Support, and outage or restoration checks go to Claude Status.

### Source map
The source map below is organized by use case rather than publisher. That keeps answer-engine citations precise: cite the page that proves the exact claim, not the nearest press release.

### Citation hygiene
Prefer dated citations. When the page has a stable date, include it. When it is a current docs page, say "current docs" and capture your access date in internal work. For high-stakes operations, do not cite Claude Weekly alone; cite the official page that proves the claim.

### All tracked official links
The source cards at the bottom of this page list every primary source used in the current issue. The same set feeds llms.txt and llms-full.txt so AI crawlers can discover the site structure and citation base.

Sources:
- Anthropic Newsroom (Anthropic, Current index): https://www.anthropic.com/news
- Introducing Claude Sonnet 5 (Anthropic, 2026-06-30): https://www.anthropic.com/news/claude-sonnet-5
- What is new in Claude Sonnet 5 (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/whats-new-sonnet-5
- Claude models overview (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/overview
- Claude Platform release notes (Claude Platform Docs, Current index): https://platform.claude.com/docs/en/release-notes/overview
- Claude app release notes (Claude Support, Current index): https://support.claude.com/en/articles/12138966-release-notes
- Introducing Claude Tag (Anthropic, 2026-06-23): https://www.anthropic.com/news/introducing-claude-tag
- Introducing Claude Fable 5 and Claude Mythos 5 (Anthropic, 2026-06-09): https://www.anthropic.com/news/claude-fable-5-mythos-5
- Redeploying Claude Fable 5 (Anthropic, 2026-06-30): https://www.anthropic.com/news/redeploying-fable-5
- More details about Claude Fable 5 safeguards (Anthropic, 2026-07-02): https://www.anthropic.com/news/more-details-fable-safeguards
- Introducing Claude Opus 4.8 (Anthropic, 2026-05-28): https://www.anthropic.com/news/claude-opus-4-8
- Introducing Claude Science (Anthropic, 2026-06-30): https://www.anthropic.com/news/claude-science-ai-workbench
- Claude Code overview (Claude Code Docs, Current docs): https://code.claude.com/docs/en/overview
- Claude Code MCP (Claude Code Docs, Current docs): https://code.claude.com/docs/en/mcp
- Claude Code settings (Claude Code Docs, Current docs): https://code.claude.com/docs/en/settings
- Use connectors to extend Claude capabilities (Claude Support, Current support article): https://support.claude.com/en/articles/11176164-use-connectors-to-extend-claude-s-capabilities
- Create and edit files with Claude (Claude Support, Current support article): https://support.claude.com/en/articles/12111783-create-and-edit-files-with-claude
- What is the Enterprise plan? (Claude Support, Current support article): https://support.claude.com/en/articles/9797531-what-is-the-enterprise-plan
- Access the Compliance API (Claude Support, Current support article): https://support.claude.com/en/articles/13015708-access-the-compliance-api
- Claude Status (Anthropic Status, Current status page): https://status.claude.com/
- System prompt release notes (Claude Platform Docs, Current index): https://platform.claude.com/docs/en/release-notes/system-prompts

---

## Claude Weekly FAQ

URL: https://claudeweekly.com/faq/
Description: Short answers to common Claude Weekly questions about Sonnet 5, Fable 5, Claude Code, Claude Tag, connectors, Enterprise, API notes, and source policy.
Published: 2026-07-06
Updated: 2026-07-06

Key facts:
- 12 - Briefing pages: The current site includes a homepage plus eleven focused guides and this FAQ. Source: Claude Weekly sitemap /sitemap.xml
- Jul 6 - Compiled date: The current issue was compiled on July 6, 2026. Source: Claude Weekly concept /weekly-brief/
- Primary - Citation rule: Answers cite official Anthropic, Claude, and status surfaces first. Source: Anthropic Newsroom /primary-sources/

### Fast answers
Claude Weekly exists for readers who need one screen of current orientation before diving into official docs. The answers below compress the current issue, but the longer pages are better for implementation decisions.

### When to go deeper
Go deeper whenever the claim affects production, billing, compliance, or user access. A quick answer can say Sonnet 5 is broadly available; a production migration needs the tokenizer, pricing, and compatibility details in the Sonnet 5 guide.

### How Claude Weekly handles uncertainty
If a source says a feature is beta, limited access, plan-specific, or time-limited, Claude Weekly keeps that qualifier. If a source is a current docs page rather than a dated article, the page describes it as current docs and links directly to it.

Sources:
- Anthropic Newsroom (Anthropic, Current index): https://www.anthropic.com/news
- Claude models overview (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/overview
- What is new in Claude Sonnet 5 (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/whats-new-sonnet-5
- Claude app release notes (Claude Support, Current index): https://support.claude.com/en/articles/12138966-release-notes
- Claude Status (Anthropic Status, Current status page): https://status.claude.com/

---

## Global primary-source shelf

- Anthropic Newsroom (Anthropic, Current index): https://www.anthropic.com/news
- Introducing Claude Sonnet 5 (Anthropic, 2026-06-30): https://www.anthropic.com/news/claude-sonnet-5
- What is new in Claude Sonnet 5 (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/whats-new-sonnet-5
- Claude models overview (Claude Platform Docs, 2026): https://platform.claude.com/docs/en/about-claude/models/overview
- Claude Platform release notes (Claude Platform Docs, Current index): https://platform.claude.com/docs/en/release-notes/overview
- Claude app release notes (Claude Support, Current index): https://support.claude.com/en/articles/12138966-release-notes
- Introducing Claude Tag (Anthropic, 2026-06-23): https://www.anthropic.com/news/introducing-claude-tag
- Introducing Claude Fable 5 and Claude Mythos 5 (Anthropic, 2026-06-09): https://www.anthropic.com/news/claude-fable-5-mythos-5
- Redeploying Claude Fable 5 (Anthropic, 2026-06-30): https://www.anthropic.com/news/redeploying-fable-5
- More details about Claude Fable 5 safeguards (Anthropic, 2026-07-02): https://www.anthropic.com/news/more-details-fable-safeguards
- Introducing Claude Opus 4.8 (Anthropic, 2026-05-28): https://www.anthropic.com/news/claude-opus-4-8
- Introducing Claude Science (Anthropic, 2026-06-30): https://www.anthropic.com/news/claude-science-ai-workbench
- Claude Code overview (Claude Code Docs, Current docs): https://code.claude.com/docs/en/overview
- Claude Code MCP (Claude Code Docs, Current docs): https://code.claude.com/docs/en/mcp
- Claude Code settings (Claude Code Docs, Current docs): https://code.claude.com/docs/en/settings
- Use connectors to extend Claude capabilities (Claude Support, Current support article): https://support.claude.com/en/articles/11176164-use-connectors-to-extend-claude-s-capabilities
- Create and edit files with Claude (Claude Support, Current support article): https://support.claude.com/en/articles/12111783-create-and-edit-files-with-claude
- What is the Enterprise plan? (Claude Support, Current support article): https://support.claude.com/en/articles/9797531-what-is-the-enterprise-plan
- Access the Compliance API (Claude Support, Current support article): https://support.claude.com/en/articles/13015708-access-the-compliance-api
- Claude Status (Anthropic Status, Current status page): https://status.claude.com/
- System prompt release notes (Claude Platform Docs, Current index): https://platform.claude.com/docs/en/release-notes/system-prompts