StackPrism

Knowledge cutoff

Claude Fable 5.1 knowledge cutoff: what it doesn’t know about your stack

Claude Fable 5.1’s reliable knowledge cutoff and training data cutoff are both June 2026, 4 months ago, per Anthropic’s docs. Anything that shipped, broke or reached end of life since then is news to it.

Reliable knowledge cutoff
June 2026
Training data cutoff
June 2026
Released
September 1, 2026
In coding agents
Claude Code, Cursor and GitHub Copilot

Source: Anthropic docs: Claude Fable 5.1 model page, checked . The terms and dates are Anthropic’s own.

What changed since June 2026

Since June 2026, CISA has added 24 actively exploited vulnerabilities in web-stack software to its Known Exploited Vulnerabilities catalog, GitHub has published 35 critical advisories for widely used npm and PyPI packages, 6 release lines have reached end of life and 5 new release lines have shipped. Claude Fable 5.1 can’t know about any of it unless a tool gives it live data.

  • 24Exploited in the wild
  • 35Critical advisories
  • 6Reached end of life
  • 5New release lines

Exploited in the wild24

The most notable, one per product or package, newest first.

  • CVE-2023-22894 added to CISA’s exploited list (Strapi)

    Strapi contains a cleartext storage of sensitive information vulnerability that could allow attackers with access to the admin panel to discover sensitive user details…

    , CISA KEV

  • CVE-2026-85706 added to CISA’s exploited list (GitLab)

    GitLab Community Edition and Enterprise Edition contains a path traversal vulnerability that allows an unauthenticated user to read arbitrary files due to an improper…

    , CISA KEV

  • CVE-2026-48710 added to CISA’s exploited list (Starlette)

    Kludex Starlette contains a HTTP request/response smuggling vulnerability that could allow attackers to inject paths into the host part, prepending the actual path…

    , CISA KEV

  • CVE-2026-34486 added to CISA’s exploited list (Apache Tomcat)

    Apache Tomcat contains a missing encryption of sensitive data vulnerability that allows the bypass of the EncryptInterceptor.

    , CISA KEV

  • CVE-2026-60137 added to CISA’s exploited list (WordPress)

    WordPress Core contains a SQL injection vulnerability when a plugin or theme passes untrusted input to the parameter.

    , CISA KEV

Show the other 19

  • CVE-2016-3081 added to CISA’s exploited list (Apache Struts)

    Apache Struts contains a command injection vulnerability that could allow remote attackers to execute arbitrary code via method:prefix when Dynamic Method Invocation is…

    , CISA KEV

  • CVE-2026-87902 added to CISA’s exploited list (WordPress)

    WordPress Core contains a remote file inclusion vulnerability which could allow an unauthenticated attacker to make page-template resolution include a chosen readable…

    , CISA KEV

  • CVE-2026-71362 added to CISA’s exploited list (Magento)

    Adobe Commerce and Magento contains an incorrect authorization vulnerability that could allow an attacker to leverage this vulnerability to gain elevated access to…

    , CISA KEV

  • CVE-2026-42016 added to CISA’s exploited list (JFrog Artifactory)

    JFrog Artifactory contains an incorrect authorization vulnerability that leads to a privilege escalation attack due to a validation check of the token signature/issuer…

    , CISA KEV

  • CVE-2026-42018 added to CISA’s exploited list (JFrog Artifactory)

    JFrog Artifactory contains an improper authentication vulnerability that could return an internal anonymous-user token to an unauthenticated caller when anonymous access…

    , CISA KEV

  • CVE-2026-75650 added to CISA’s exploited list (Magento)

    Adobe Commerce and Magento Open Source contain an improper neutralization of special elements used in a template engine vulnerability that could allow an attacker to…

    , CISA KEV

  • CVE-2026-59822 added to CISA’s exploited list (LiteLLM)

    BerriAI LiteLLM contains an improper authentication vulnerability in the MCP Streamable HTTP endpoint that could allow an unauthenticated attacker to establish an…

    , CISA KEV

  • CVE-2026-49869 added to CISA’s exploited list (Kestra)

    Kestra OSS contains an OS command injection vulnerability that could allow an unauthenticated remote attacker to create and execute arbitrary workflows without…

    , CISA KEV

  • CVE-2026-82329 added to CISA’s exploited list (JFrog Artifactory)

    JFrog Artifactory contains an improper authentication vulnerability that under default configuration can allow an unauthenticated attacker with network access to obtain…

    , CISA KEV

  • CVE-2026-66384 added to CISA’s exploited list (JFrog Artifactory)

    JFrog Artifactory contains an improper limitation of a pathname to a restricted directory vulnerability.

    , CISA KEV

  • CVE-2026-60004 added to CISA’s exploited list (Gitea)

    Gitea contains a code injection vulnerability that allows an attacker with repository write access to send a malicious patch to the diffpatch API endpoint to plant an…

    , CISA KEV

  • CVE-2026-64849 added to CISA’s exploited list (MLflow)

    MLflow contains a server-side request forgery vulnerability that can allow attackers to reach internal or cloud metadata services and receive response_status and…

    , CISA KEV

  • CVE-2025-62593 added to CISA’s exploited list (Ray)

    Ray-Project Ray contains a code injection vulnerability that could allow remote code execution.

    , CISA KEV

  • CVE-2026-72898 added to CISA’s exploited list (Metabase)

    Metabase contains a SQL Injection vulnerability that allows an unauthenticated remote attacker to inject arbitrary SQL into the Metabase application database, which can…

    , CISA KEV

  • CVE-2026-63077 added to CISA’s exploited list (TeamCity)

    JetBrains TeamCity contains a deserialization of untrusted data vulnerability that could allow unauthenticated remote code execution via the agent polling protocol.

    , CISA KEV

  • CVE-2026-9198 added to CISA’s exploited list (Langflow)

    Langflow contains a code injection vulnerability that allows unauthenticated attackers to achieve full remote code execution on default Langflow deployments.

    , CISA KEV

  • CVE-2026-63030 added to CISA’s exploited list (WordPress)

    WordPress Core contains an interpretation conflict vulnerability that could allow an attacker to perform SQL Injection and achieve Remote Code Execution.

    , CISA KEV

  • CVE-2026-0770 added to CISA’s exploited list (Langflow)

    Langflow contains an inclusion of functionality from untrusted control sphere vulnerability that allows remote attackers to execute arbitrary code on affected…

    , CISA KEV

  • CVE-2026-55255 added to CISA’s exploited list (Langflow)

    Langflow contains an authorization bypass through user-controlled key vulnerability which allows an authenticated attacker to execute any flow belonging to another user…

    , CISA KEV

Critical advisories35

The most notable, one per product or package, newest first.

Show the other 30

  • handlebars: JavaScript injection via Own Property check bypass (critical)

    Fixed in 4.7.10.

    , GitHub Advisory Database

  • payload: unauthorized update to collection documents (critical)

    Fixed in 3.90.0.

    , GitHub Advisory Database

  • payload: Payload authentication token field handling issue (critical)

    Fixed in 3.90.0.

    , GitHub Advisory Database

  • payload: field access control bypass on auth collections (critical)

    Fixed in 3.90.0.

    , GitHub Advisory Database

  • payload: SQL injection in SQLite and Postgres (critical)

    Fixed in 3.88.0.

    , GitHub Advisory Database

  • langflow: PythonREPLComponent executes unsandboxed Python code, enabling… (critical)

    Fixed in 1.10.1.

    , GitHub Advisory Database

  • simple-git: simple-git unsafe-operation guard does not block trailer command… (critical)

    Fixed in 4.0.1.

    , GitHub Advisory Database

  • seroval: fromJSON() Promise thenable assimilation invokes plugin-produced… (critical)

    Fixed in 1.6.2.

    , GitHub Advisory Database

  • proxy-addr: IP spoofing via IPv4-mapped IPv6 trust subnet (critical)

    Fixed in 2.0.8.

    , GitHub Advisory Database

  • langflow: weak Fernet key via random.seed() (critical)

    Fixed in 1.10.1.

    , GitHub Advisory Database

  • langflow: unauthenticated Flow execution via webhook authentication bypass (critical)

    Fixed in 1.9.1.

    , GitHub Advisory Database

  • pyjwt: Asymmetric-PEM detection bypass: whitespace/line-ending-mutated public… (critical)

    Fixed in 2.14.0.

    , GitHub Advisory Database

  • anyio: TLSStream IDNA 2003 host name encoding enables potential TLS… (critical)

    Fixed in 4.14.2.

    , GitHub Advisory Database

  • astro: remote code execution through AVIF image optimization (critical)

    Fixed in 7.2.8.

    , GitHub Advisory Database

  • next: unauthenticated remote code execution in image optimization API (critical)

    Fixed in 15.5.24 and 16.3.3.

    , GitHub Advisory Database

  • next: unauthenticated remote code execution on windows-hosted servers (critical)

    Fixed in 15.5.24 and 16.3.3.

    , GitHub Advisory Database

  • gitpython: Dormant multi-line git-config values are corrupted into live injected… (critical)

    Fixed in 3.1.59.

    , GitHub Advisory Database

  • nltk: Allowlisted pickle loaders still permit code execution in current source (critical)

    Fixed in 3.10.3.

    , GitHub Advisory Database

  • nltk: unsafe Pickle deserialization in TransitionParser Allows remote code… (critical)

    Fixed in 3.10.0.

    , GitHub Advisory Database

  • python-jose: python-jose algorithm confusion guard bypassed (critical)

    , GitHub Advisory Database

  • nltk: JVM argument injection bypass (critical)

    Fixed in 3.10.3.

    , GitHub Advisory Database

  • chainlit: command injection (critical)

    Fixed in 2.12.0.

    , GitHub Advisory Database

  • crypto-js: Insufficient Entropy in Cryptographic secret Generation (critical)

    Fixed in 4.0.0.

    , GitHub Advisory Database

  • @nuxt/devtools: unauthenticated Nuxt DevTools RPC allows arbitrary command execution on… (critical)

    Fixed in 3.3.1.

    , GitHub Advisory Database

  • sequelize: SQL injection (Oracle DB) (critical)

    Fixed in 6.37.4.

    , GitHub Advisory Database

  • sentence-transformers: sentence-transformers local model loading bypasses trust_remote_code… (critical)

    Fixed in 5.6.0.

    , GitHub Advisory Database

  • seroval: seroval.fromJSON() Promise resolver type confusion invokes… (critical)

    Fixed in 1.5.3.

    , GitHub Advisory Database

  • @auth/core: email normalizer validates the address before Unicode normalization (critical)

    Fixed in 0.41.3.

    , GitHub Advisory Database

  • tar: Decompression/parse DoS via unlimited input (critical)

    Fixed in 7.5.19.

    , GitHub Advisory Database

  • better-auth: OAuth refresh-token replay (critical)

    Fixed in 1.6.11.

    , GitHub Advisory Database

Reached end of life6

The most notable, one per product or package, newest first.

  • Python 3.10 reached end of life

    No more security fixes. Latest: Python 3.15.

    , endoflife.date

  • Debian 11 reached end of life

    No more security fixes. Latest: Debian 13.

    , endoflife.date

  • Go 1.25 reached end of life

    No more security fixes. Latest: Go 1.27.

    , endoflife.date

  • MongoDB 8.2 reached end of life

    No more security fixes. Latest: MongoDB 9.0.

    , endoflife.date

  • Nuxt 3 reached end of life

    No more security fixes. Latest: Nuxt 4.

    , endoflife.date

Show the other 1

  • Ruby on Rails 7.2 reached end of life

    No more security fixes. Latest: Ruby on Rails 8.1.

    , endoflife.date

New release lines5

  • Python 3.15 released

    Latest version: 3.15.0.

    , endoflife.date

  • MongoDB 9.0 released

    Latest version: 9.0.4.

    , endoflife.date

  • Kubernetes 1.37 released

    Latest version: 1.37.1.

    , endoflife.date

  • Go 1.27 released

    Latest version: 1.27.2.

    , endoflife.date

  • Django 6.1 released

    Latest version: 6.1.2.

    , endoflife.date

Events last checked against CISA KEV, GitHub Advisory Database and endoflife.date.

What counts: vulnerabilities CISA confirms are exploited, in software web apps commonly run; critical GitHub advisories on widely used npm and PyPI packages; and release lines of popular runtimes, frameworks and databases reaching end of life or shipping.

How to give Claude Code, Cursor or GitHub Copilot today’s data

Claude Fable 5.1 only knows what was in its training data. StackPrism’s MCP server lets your agent check packages and services against live vulnerability, end-of-life and outage data while it works. How the MCP server works.

  1. Create a free account and an access token on the Connect page.
  2. Add the server to your agent, with your token in place of <YOUR_TOKEN>.
  3. Ask it to “check whether my dependencies are safe with StackPrism”.
Claude Code

Run this once in a terminal. Add --scope user to use StackPrism in every project, not just this folder.

Add the server

claude mcp add --transport http stackprism https://stackprism.ai/api/mcp --header "Authorization: Bearer <YOUR_TOKEN>"

Check it with claude mcp list, or /mcp inside a session.

Cursor

Add this to .cursor/mcp.json in your project, or ~/.cursor/mcp.json for every project. Keep the file out of git: it holds your token.

mcp.json

{
  "mcpServers": {
    "stackprism": {
      "url": "https://stackprism.ai/api/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_TOKEN>"
      }
    }
  }
}

StackPrism then shows under MCP in Cursor’s settings.

GitHub Copilot

Any MCP client that supports Streamable HTTP can connect with the URL and header below. For clients that only run local servers, use the mcp-remote bridge.

Server URL

https://stackprism.ai/api/mcp

Header

Authorization: Bearer <YOUR_TOKEN>

Or bridge it over stdio (needs Node)

npx -y mcp-remote https://stackprism.ai/api/mcp --header "Authorization: Bearer <YOUR_TOKEN>"

Check your repo against today

Paste a public GitHub repo and see which dependencies are vulnerable, end of life or deprecated today, with a fix task to hand your coding agent.

Free, no sign-up. No repo handy? See a sample report.

Questions

What is Claude Fable 5.1’s knowledge cutoff?

Claude Fable 5.1’s reliable knowledge cutoff and training data cutoff are both June 2026, 4 months ago, per Anthropic’s docs. Source: Claude Fable 5.1 model page.

What’s the difference between its reliable knowledge cutoff and training data cutoff?

Anthropic defines the reliable knowledge cutoff as “the date through which the model’s knowledge is most extensive and reliable”. The training data cutoff is the broader range of data used to train it, so material from between the two dates is in there, but less of it.

What has changed since Claude Fable 5.1’s cutoff?

Since June 2026, CISA has added 24 actively exploited vulnerabilities in web-stack software to its Known Exploited Vulnerabilities catalog, GitHub has published 35 critical advisories for widely used npm and PyPI packages, 6 release lines have reached end of life and 5 new release lines have shipped.

The most notable: CVE-2026-48710 added to CISA’s exploited list (Starlette) (Sep 2, 2026); handlebars: JavaScript injection (critical) (Oct 8, 2026); Python 3.10 reached end of life (Oct 1, 2026).

How do I give Claude Fable 5.1 current information?

Give it a tool that reads live data. StackPrism’s MCP server lets your coding agent check packages and services against live vulnerability, end-of-life and outage data while it works. Set it up in a few minutes.