# MSP MCP

MSP MCP is a family of Model Context Protocol servers that expose Managed Service Provider (MSP) tooling to AI assistants. Each server wraps a vendor's API -- PSAs (Autotask, ConnectWise, HaloPSA), RMMs (Datto, NinjaOne, Atera, Syncro), documentation platforms (IT Glue, Hudu, Liongard), security tools (SentinelOne, Huntress, Blumira, RocketCyber), email security (Proofpoint, Avanan, Abnormal, Mimecast, IRONSCALES, KnowBe4, SpamTitan), incident management (Rootly, PagerDuty), and more -- as MCP-compliant tools so an MSP technician can ask Claude to triage a ticket, check device health, or pull documentation in plain English.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## How to limit MCP tool output size in Claude Code

DevFeed: [How to limit MCP tool output size in Claude Code](<https://devfeed.tech/articles/how-to-limit-mcp-tool-output-size-in-claude-code-144281.md>)

Original publisher: [Read original article](<https://nirmata.com/2026/11/20/how-to-limit-mcp-tool-output-size-in-claude-code/>)

Author: Ritesh Patel

Published: 2026-11-19T20:41:35.692178Z

Content type: tutorial

Language: en

Sources: [Nirmata](<https://devfeed.tech/sources/nirmata.md>)

Topics: [Claude Code](<https://devfeed.tech/topics/claude-code.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [Claude](<https://devfeed.tech/topics/claude.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [airtable](<https://devfeed.tech/tags/airtable.md>), [audit](<https://devfeed.tech/tags/audit.md>), [call](<https://devfeed.tech/tags/call.md>), [claude](<https://devfeed.tech/tags/claude.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [context](<https://devfeed.tech/tags/context.md>), [control](<https://devfeed.tech/tags/control.md>), [cost](<https://devfeed.tech/tags/cost.md>), [cost-management](<https://devfeed.tech/tags/cost-management.md>), [file](<https://devfeed.tech/tags/file.md>), [how-to](<https://devfeed.tech/tags/how-to.md>), [limit](<https://devfeed.tech/tags/limit.md>), [max](<https://devfeed.tech/tags/max.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-tool](<https://devfeed.tech/tags/mcp-tool.md>), [security](<https://devfeed.tech/tags/security.md>), [set](<https://devfeed.tech/tags/set.md>), [size](<https://devfeed.tech/tags/size.md>), [text](<https://devfeed.tech/tags/text.md>), [tokens](<https://devfeed.tech/tags/tokens.md>), [tool](<https://devfeed.tech/tags/tool.md>)

### AI overview

Set `MAX_MCP_OUTPUT_TOKENS` in the shell before starting Claude Code to control when successful MCP results are saved to the session's `tool-results` directory and replaced in the conversation by a file path. Its documented default is 25,000 tokens; a separate 50,000-character threshold for text results remains fixed, and the 10,000-token transcript warning does not stop output. Saving a result to disk does not prevent Claude from reading it later. The source also distinguishes server startup timeout (`MCP_TIMEOUT`, default 30,000 ms) from tool execution timeout (`MCP_TOOL_TIMEOUT`, default 100,000,000 ms), which a server's `.mcp.json` `timeout` can override. The token setting does not filter server output, establish server trust, or change transport limits. Server developers can raise a tool's text threshold with `anthropic/maxResultSizeChars`, up to 500,000 characters.

### Source excerpt

Set MAX_MCP_OUTPUT_TOKENS before starting Claude Code. The default is 25,000 tokens, and text over 50,000 characters is saved to a file whatever the setting.

## Your Coding Agent Can Read the Code. Give It the History with Relate by Neo4j

DevFeed: [Your Coding Agent Can Read the Code. Give It the History with Relate by Neo4j](<https://devfeed.tech/articles/your-coding-agent-can-read-the-code-give-it-the-history-with-relate-by-neo4j-144105.md>)

Original publisher: [Read original article](<https://neo4j.com/blog/developer/your-coding-agent-can-read-the-code-give-it-the-history-with-relate-by-neo4j/>)

Author: Kate Smith

Published: 2026-11-18T21:00:00Z

Content type: tutorial

Language: en

Sources: [Neo4j](<https://devfeed.tech/sources/graph-database-technology-neo4j-blog.md>)

Topics: [Neo4j](<https://devfeed.tech/topics/neo4j.md>), [code search](<https://devfeed.tech/topics/code-search.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>), [codex](<https://devfeed.tech/topics/codex.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agent-skill](<https://devfeed.tech/tags/agent-skill.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [api-key](<https://devfeed.tech/tags/api-key.md>), [cli](<https://devfeed.tech/tags/cli.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agent](<https://devfeed.tech/tags/coding-agent.md>), [developer](<https://devfeed.tech/tags/developer.md>), [extension](<https://devfeed.tech/tags/extension.md>), [git](<https://devfeed.tech/tags/git.md>), [graph](<https://devfeed.tech/tags/graph.md>), [graph-database](<https://devfeed.tech/tags/graph-database.md>), [history](<https://devfeed.tech/tags/history.md>), [load](<https://devfeed.tech/tags/load.md>), [neo4j](<https://devfeed.tech/tags/neo4j.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [relate](<https://devfeed.tech/tags/relate.md>), [repository](<https://devfeed.tech/tags/repository.md>), [security](<https://devfeed.tech/tags/security.md>), [source](<https://devfeed.tech/tags/source.md>)

### AI overview

Relate combines a repository's Git history and working tree in one graph, letting coding agents find co-changed files, prior change patterns, and contributors that source code alone may not reveal. In the FastAPI example, `api_key.py` repeatedly changed with three security modules that do not reference it; all three tested agents found those siblings with the graph, while none did from source code alone. The post shows importing history with `git-to-graph` and file text with `fs-to-graph`, then querying the store through read-only MCP or the CLI. The comparison also found that agents could identify tests and imported files from code, but needed history to identify relevant past contributors. File-text matching for unreferenced pairs is approximate, and `fs-to-graph` defaults to a 64 KB content cap, which can omit large files unless raised.

### Source excerpt

Load a repository's git history into a graph with Relate's git-to-graph extension, serve it read-only, and see what a coding agent finds with it that it cannot find in the source tree. In a popular open source codebase, security/api_key.py has... Read more ->

## Claude-Mem Tested: Perfect Recall, but at a Cost of 5,200 Tokens Per Request

DevFeed: [Claude-Mem Tested: Perfect Recall, but at a Cost of 5,200 Tokens Per Request](<https://devfeed.tech/articles/claude-mem-tested-perfect-recall-but-at-a-cost-of-5-200-tokens-per-request-140468.md>)

Original publisher: [Read original article](<https://hackernoon.com/claude-mem-tested-perfect-recall-but-at-a-cost-of-5200-tokens-per-request?source=rss>)

Author: Ishank Choudhary

Published: 2026-11-18T05:56:23Z

Content type: article

Language: en

Sources: [HackerNoon](<https://devfeed.tech/sources/hackernoon.md>)

Topics: [AI chatbot continuity](<https://devfeed.tech/topics/ai-chatbot-continuity.md>), [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [Claude](<https://devfeed.tech/topics/claude.md>), [API](<https://devfeed.tech/topics/api.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>), [Test-driven development](<https://devfeed.tech/topics/tdd.md>), [API keys](<https://devfeed.tech/topics/api-keys.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [2025](<https://devfeed.tech/tags/2025.md>), [2026](<https://devfeed.tech/tags/2026.md>), [3](<https://devfeed.tech/tags/3.md>), [about](<https://devfeed.tech/tags/about.md>), [added](<https://devfeed.tech/tags/added.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-memory-persistence](<https://devfeed.tech/tags/ai-memory-persistence.md>), [but](<https://devfeed.tech/tags/but.md>), [capture](<https://devfeed.tech/tags/capture.md>), [claude](<https://devfeed.tech/tags/claude.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [claude-mem-vs-no-memory](<https://devfeed.tech/tags/claude-mem-vs-no-memory.md>), [context-engineering](<https://devfeed.tech/tags/context-engineering.md>), [cost](<https://devfeed.tech/tags/cost.md>), [cross-session-memory](<https://devfeed.tech/tags/cross-session-memory.md>), [decision](<https://devfeed.tech/tags/decision.md>), [developer-tools](<https://devfeed.tech/tags/developer-tools.md>), [extra](<https://devfeed.tech/tags/extra.md>), [i](<https://devfeed.tech/tags/i.md>), [llms](<https://devfeed.tech/tags/llms.md>), [privacy](<https://devfeed.tech/tags/privacy.md>), [probabilistic](<https://devfeed.tech/tags/probabilistic.md>), [recall](<https://devfeed.tech/tags/recall.md>), [request](<https://devfeed.tech/tags/request.md>), [session](<https://devfeed.tech/tags/session.md>), [times](<https://devfeed.tech/tags/times.md>), [tokens](<https://devfeed.tech/tags/tokens.md>)

### AI overview

In a three-run test on a small invoicing repo, claude-mem carried a prior-session decision into every fresh session, enabling the correct feature 3/3 times; without memory the agent completed it 0/3 times, while re-explaining the rule restored 3/3 success. The plugin adds about 5,200 tokens per request, and full runs processed more context tokens than re-explaining the rule, with similar wall time. Its hooks send tool activity to a local worker, a second model captures observations, and later sessions receive an index and can search for details. Capture is probabilistic: explicitly use `save_memory` for critical facts and verify them. In the tested setup, free-tier Gemini quota limits and transient errors paused or dropped capture. Privacy findings: `<private>` kept tagged text out of claude-mem storage but not the agent's own history; telemetry and screenshot retention are on by default, auto-redaction is off, and the local database is readable by processes running as the user. Results come from a deliberately tiny repo, so the author says the overhead and saved exploration balance may differ on larger codebases.

### Source excerpt

claude-mem recalled a prior-session decision 3/3 times (0/3 without), at ~5,200 extra tokens/request; I tested probabilistic capture and privacy defaults.

## How to pin MCP server versions for Claude Code and Cursor

DevFeed: [How to pin MCP server versions for Claude Code and Cursor](<https://devfeed.tech/articles/how-to-pin-mcp-server-versions-for-claude-code-and-cursor-139212.md>)

Original publisher: [Read original article](<https://nirmata.com/2026/11/18/how-to-pin-mcp-server-versions-for-claude-code-and-cursor/>)

Author: Ritesh Patel

Published: 2026-11-18T01:14:41.413057Z

Content type: tutorial

Language: en

Sources: [Nirmata](<https://devfeed.tech/sources/nirmata.md>)

Topics: [versioning](<https://devfeed.tech/topics/versioning.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [Claude Code](<https://devfeed.tech/topics/claude-code.md>), [Documentation](<https://devfeed.tech/topics/documentation.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [cache](<https://devfeed.tech/tags/cache.md>), [checklist](<https://devfeed.tech/tags/checklist.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [client](<https://devfeed.tech/tags/client.md>), [command](<https://devfeed.tech/tags/command.md>), [config](<https://devfeed.tech/tags/config.md>), [credentials](<https://devfeed.tech/tags/credentials.md>), [cursor](<https://devfeed.tech/tags/cursor.md>), [dependabot](<https://devfeed.tech/tags/dependabot.md>), [dependencies](<https://devfeed.tech/tags/dependencies.md>), [docs](<https://devfeed.tech/tags/docs.md>), [documentation](<https://devfeed.tech/tags/documentation.md>), [environment-variables](<https://devfeed.tech/tags/environment-variables.md>), [launch](<https://devfeed.tech/tags/launch.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-server](<https://devfeed.tech/tags/mcp-server.md>), [pin](<https://devfeed.tech/tags/pin.md>), [security](<https://devfeed.tech/tags/security.md>), [server](<https://devfeed.tech/tags/server.md>), [supply-chain-security](<https://devfeed.tech/tags/supply-chain-security.md>), [updates](<https://devfeed.tech/tags/updates.md>), [version](<https://devfeed.tech/tags/version.md>)

### AI overview

Neither Claude Code nor Cursor documents a server-version field, range, or lockfile in the MCP configuration described. For npm servers, the article says an exact version can be named in the `npx` package specifier; that is npm behavior, and the examples were not tested in either client. It recommends recording the reviewed version alongside the config, while noting this is for reviewers and is not read by the tools. Cursor's FAQ says to remove an npm server, clear the npm cache, and re-add it to get the latest version; custom servers are updated by changing local files and restarting Cursor. Claude Code separately refreshes tool, prompt, and resource lists after `list_changed`; its `MCP_SDK_GENERATION` setting selects the client runtime, not a server version. The excerpt also advises using environment-variable references for credentials and checking tool lists after changes.

### Source excerpt

Neither Claude Code nor Cursor documents a setting to pin an MCP server version. Here is what their docs do say about launch commands, env vars and updates.

## Claude as a Teammate: A Practical Guide to AI-Powered Marketing Workflows

DevFeed: [Claude as a Teammate: A Practical Guide to AI-Powered Marketing Workflows](<https://devfeed.tech/articles/claude-as-a-teammate-a-practical-guide-to-ai-powered-marketing-workflows-138688.md>)

Original publisher: [Read original article](<https://hackernoon.com/claude-as-a-teammate-a-practical-guide-to-ai-powered-marketing-workflows?source=rss>)

Author: Daryna

Published: 2026-11-17T17:22:33Z

Content type: tutorial

Language: en

Sources: [HackerNoon](<https://devfeed.tech/sources/hackernoon.md>)

Topics: [Claude](<https://devfeed.tech/topics/claude.md>), [Claude Code](<https://devfeed.tech/topics/claude-code.md>), [Agent Skills](<https://devfeed.tech/topics/agent-skills.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [Search engine optimization (SEO)](<https://devfeed.tech/topics/seo.md>), [Website](<https://devfeed.tech/topics/website.md>), [human in the loop agent](<https://devfeed.tech/topics/human-in-the-loop-agent.md>), [Slack](<https://devfeed.tech/topics/slack.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-for-marketing](<https://devfeed.tech/tags/ai-for-marketing.md>), [ai-in-marketing](<https://devfeed.tech/tags/ai-in-marketing.md>), [articles](<https://devfeed.tech/tags/articles.md>), [b2b](<https://devfeed.tech/tags/b2b.md>), [case](<https://devfeed.tech/tags/case.md>), [case-studies](<https://devfeed.tech/tags/case-studies.md>), [claude](<https://devfeed.tech/tags/claude.md>), [claude-cowork](<https://devfeed.tech/tags/claude-cowork.md>), [connectors](<https://devfeed.tech/tags/connectors.md>), [cowork](<https://devfeed.tech/tags/cowork.md>), [engineers](<https://devfeed.tech/tags/engineers.md>), [enterprise-ai](<https://devfeed.tech/tags/enterprise-ai.md>), [generative-ai](<https://devfeed.tech/tags/generative-ai.md>), [good-company](<https://devfeed.tech/tags/good-company.md>), [guide](<https://devfeed.tech/tags/guide.md>), [marketing](<https://devfeed.tech/tags/marketing.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [pages](<https://devfeed.tech/tags/pages.md>), [seo](<https://devfeed.tech/tags/seo.md>), [skills](<https://devfeed.tech/tags/skills.md>), [team](<https://devfeed.tech/tags/team.md>), [uses](<https://devfeed.tech/tags/uses.md>), [waiting](<https://devfeed.tech/tags/waiting.md>), [workflows](<https://devfeed.tech/tags/workflows.md>)

### AI overview

BotsCrew describes using Claude as a shared marketing teammate connected to company systems, with shared instructions and repeatable playbooks. Five workflows turn records from Jira, Confluence and Fathom into reviewed case studies; use Claude Design and Claude Code to build website pages; use Ahrefs via MCP to research AI-search articles; create a Gmail domain-highlighting extension; and schedule a morning digest from email and Slack. The article reports case studies shrinking from 1-2 weeks to about 2 hours, custom pages from weeks to 2-3 hours, blog posts from 2-3 hours to about 30 minutes, and link-status checks from 1-2 hours daily to zero. It emphasizes checking outputs and sources, human approval before publishing or sending, and technical review of website code and deployment. It says the approach fits recurring internal work with accessible data, and is unsuitable as a substitute for building a customer-facing product.

### Source excerpt

How a B2B marketing team uses Claude Cowork, skills and connectors to ship case studies, web pages and SEO articles without waiting on engineers.

## Do Rule-Based Tools Help LLMs Translate Endangered Uralic Languages in the Prompt or over MCP?

DevFeed: [Do Rule-Based Tools Help LLMs Translate Endangered Uralic Languages in the Prompt or over MCP?](<https://devfeed.tech/articles/do-rule-based-tools-help-llms-translate-endangered-uralic-languages-in-the-prompt-or-over-mcp-137448.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.19439>)

Author: Iaroslav Chelombitko, Mika H\\"am\\"al\\"ainen

Published: 2026-11-17T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [machine-translation](<https://devfeed.tech/topics/machine-translation.md>), [prompt](<https://devfeed.tech/topics/prompt.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>)

Tags: [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cl](<https://devfeed.tech/tags/cs-cl.md>), [language-models](<https://devfeed.tech/tags/language-models.md>), [languages](<https://devfeed.tech/tags/languages.md>), [llms](<https://devfeed.tech/tags/llms.md>), [machine-translation](<https://devfeed.tech/tags/machine-translation.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [model-context-protocol](<https://devfeed.tech/tags/model-context-protocol.md>), [model-context-protocol-mcp](<https://devfeed.tech/tags/model-context-protocol-mcp.md>), [prompt](<https://devfeed.tech/tags/prompt.md>), [tool-calling](<https://devfeed.tech/tags/tool-calling.md>), [tools](<https://devfeed.tech/tags/tools.md>), [translation](<https://devfeed.tech/tags/translation.md>), [translations](<https://devfeed.tech/tags/translations.md>)

### AI overview

Adding rule-based resources improved translation acceptance in the expert-reviewed evaluation: accepted outputs rose from 13 to 43 of 272 when analyses were placed in the prompt, and to 55 with tools attached. The reported gain came from bilingual dictionary glosses. For Udmurt, whose dictionary lacks an English side, prompt-based access produced no gain; for the other three languages, English glosses covered 74-80% of source words. MCP tool use was harder to make work: 14 of 31 open-weight models never called a tool, including 11 unable to produce a usable call, and half of calls named unsupported languages. The evaluation covered 816 sentences across three conditions, and the authors say they release the dataset, outputs, and call logs.

### Source excerpt

arXiv:2611.19439v1 Announce Type: cross Abstract: Komi-Zyrian, Moksha, Erzya and Udmurt have mature rule-based resources, among them finite-state analysers, constraint grammars and bilingual dictionaries, which language models do not use. We ask whether giving a model access to these resources improves machine translation from the four languages into English, and whether it matters how the access is provided: as analyses computed in advance and written into the prompt, or as tools that the model calls itself through the Model Context Protocol (MCP). We evaluate manually: an expert linguist read every English output of the strongest model under three conditions, 816 sentences in all, accepted or rejected each one, and named the main error in every rejected translation. With the analyses in the prompt, the number of accepted translations rises from 13 to 43 out of 272; with the tools attached it rises to 55. The gain comes from the bilingual dictionary. For Udmurt, whose dictionary has no English side, the prompt-based condition brings no gain at all, while for the other three languages an English gloss reaches the prompt for 74 to 80% of the source words. Tool calling is also much harder to get working than prompt injection. Of 31 open-weight models, 14 never call a tool, 11 of them because they cannot produce a usable call, and half of the calls that are made name a language the toolkit does not support. We release the data set, every model output and all tool-call logs.

## Introducing Insights Agent

DevFeed: [Introducing Insights Agent](<https://devfeed.tech/articles/introducing-insights-agent-139723.md>)

Original publisher: [Read original article](<https://planetscale.com/blog/introducing-insights-agent>)

Author: Enrique Gonzalez

Published: 2026-11-17T00:00:00Z

Content type: release

Language: en

Sources: [PlanetScale](<https://devfeed.tech/sources/blog-planetscale.md>)

Topics: [AI Agent](<https://devfeed.tech/topics/ai-agent.md>), [Pull Request](<https://devfeed.tech/topics/pull-request.md>), [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [agent observability](<https://devfeed.tech/topics/agent-observability.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [anthropic](<https://devfeed.tech/tags/anthropic.md>), [cli](<https://devfeed.tech/tags/cli.md>), [cloud-agents](<https://devfeed.tech/tags/cloud-agents.md>), [code](<https://devfeed.tech/tags/code.md>), [connect](<https://devfeed.tech/tags/connect.md>), [cursor-origin](<https://devfeed.tech/tags/cursor-origin.md>), [insights](<https://devfeed.tech/tags/insights.md>), [product](<https://devfeed.tech/tags/product.md>), [pull-requests](<https://devfeed.tech/tags/pull-requests.md>), [queries](<https://devfeed.tech/tags/queries.md>), [recommendations](<https://devfeed.tech/tags/recommendations.md>), [schema](<https://devfeed.tech/tags/schema.md>)

### AI overview

PlanetScale Insights Agent investigates database performance issues using a database's schema, Query Insights data, and connected repository, then opens a pull request when evidence supports a change. Its two focused run types address slow queries and recommendations; runs can be started on demand or scheduled, and may propose changes across multiple files. It follows repository conventions, and Vitess runs can also open a deploy request. Changes are suggestions: nothing is automatically merged or deployed. Runs are billed when completed, with organization spend limits available; the temporary sandbox exists only during a run. PlanetScale's built-in agent offers curated prompts and managed setup, while the CLI and MCP server let teams build their own automations with custom prompts and more control. At launch, one GitHub or Cursor Origin repository can connect to each database. The feature is on by default but acts only after a repository is connected.

### Source excerpt

Insights Agent connect to your codebase to turn slow queries and schema recommendations into pull requests.

## Building a Secure MCP (Model Context Protocol) Server with Auth0

DevFeed: [Building a Secure MCP (Model Context Protocol) Server with Auth0](<https://devfeed.tech/articles/building-a-secure-mcp-model-context-protocol-server-with-auth0-136615.md>)

Original publisher: [Read original article](<https://auth0.com/blog/building-secure-mcp-server-with-auth0/>)

Author: Kevin Kimani

Published: 2026-11-16T17:06:00Z

Content type: tutorial

Language: en

Sources: [Auth0 Blog](<https://devfeed.tech/sources/auth0-blog.md>)

Topics: [Auth0](<https://devfeed.tech/topics/auth0.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [Secure token management](<https://devfeed.tech/topics/secure-token-management.md>), [Single sign-on (SSO)](<https://devfeed.tech/topics/sso.md>), [Back end](<https://devfeed.tech/topics/backend.md>), [Test-driven development](<https://devfeed.tech/topics/tdd.md>), [Remote Procedure Call (RPC)](<https://devfeed.tech/topics/rpc.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [auth0](<https://devfeed.tech/tags/auth0.md>), [build](<https://devfeed.tech/tags/build.md>), [building](<https://devfeed.tech/tags/building.md>), [context](<https://devfeed.tech/tags/context.md>), [data](<https://devfeed.tech/tags/data.md>), [go](<https://devfeed.tech/tags/go.md>), [how-to](<https://devfeed.tech/tags/how-to.md>), [learn](<https://devfeed.tech/tags/learn.md>), [learn-how-to](<https://devfeed.tech/tags/learn-how-to.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-server](<https://devfeed.tech/tags/mcp-server.md>), [model](<https://devfeed.tech/tags/model.md>), [model-context-protocol](<https://devfeed.tech/tags/model-context-protocol.md>), [protocol](<https://devfeed.tech/tags/protocol.md>), [server](<https://devfeed.tech/tags/server.md>), [with](<https://devfeed.tech/tags/with.md>)

### AI overview

The guide builds a Go MCP server that uses Auth0 Authorization Code with PKCE to issue user-linked access tokens, then validates JWT signatures, issuer, audience, and expiry against Auth0 JWKS. Auth0 RBAC supplies department permissions; the server authorizes from the token's requested-and-approved `scope` claim, mapping `read:engineering` or `read:hr` to a department. It registers only that department's resources for each session and returns "resource not found" for cross-department reads to reduce information leakage. Missing or invalid tokens receive 401 with OAuth resource metadata; valid tokens without a required scope receive 403. The walkthrough also configures OAuth Protected Resource Metadata and demonstrates requests with cURL. It distinguishes Client Credentials as suitable for automated machine-to-machine pipelines, while choosing Authorization Code for deployments governed by a real user. Production follow-ups include token refresh, client-secret rotation, richer audit logging, and finer-grained authorization for more complex access patterns.

### Source excerpt

Learn how to build an MCP server in Go and secure it with Auth0 using OAuth 2.1, PKCE, and RBAC-backed scopes, so AI agents only see the data their user is authorized to access.

## Security Journey brings training program management into Claude and ChatGPT

DevFeed: [Security Journey brings training program management into Claude and ChatGPT](<https://devfeed.tech/articles/security-journey-brings-training-program-management-into-claude-and-chatgpt-136880.md>)

Original publisher: [Read original article](<https://sdtimes.com/ai-assistants/security-journey-brings-training-program-management-into-claude-and-chatgpt/>)

Author: David Lyman

Published: 2026-11-16T16:57:50Z

Content type: news

Language: en

Sources: [SD Times](<https://devfeed.tech/sources/sd-times.md>)

Topics: [AI Integration Strategies](<https://devfeed.tech/topics/ai-integration-strategies.md>), [Claude](<https://devfeed.tech/topics/claude.md>), [ChatGPT](<https://devfeed.tech/topics/chatgpt.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-assistants](<https://devfeed.tech/tags/ai-assistants.md>), [chatgpt](<https://devfeed.tech/tags/chatgpt.md>), [claude](<https://devfeed.tech/tags/claude.md>), [code](<https://devfeed.tech/tags/code.md>), [integrations](<https://devfeed.tech/tags/integrations.md>), [journey](<https://devfeed.tech/tags/journey.md>), [language](<https://devfeed.tech/tags/language.md>), [latest-news](<https://devfeed.tech/tags/latest-news.md>), [management](<https://devfeed.tech/tags/management.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [model-context-protocol](<https://devfeed.tech/tags/model-context-protocol.md>), [model-context-protocol-mcp](<https://devfeed.tech/tags/model-context-protocol-mcp.md>), [program](<https://devfeed.tech/tags/program.md>), [program-management](<https://devfeed.tech/tags/program-management.md>), [reports](<https://devfeed.tech/tags/reports.md>), [secure-coding-training](<https://devfeed.tech/tags/secure-coding-training.md>), [security](<https://devfeed.tech/tags/security.md>), [security-journey](<https://devfeed.tech/tags/security-journey.md>), [training](<https://devfeed.tech/tags/training.md>)

### AI overview

Security Journey's Claude and ChatGPT integrations let program owners manage training in natural language without leaving those assistants. They can find learners who are behind, assign training, answer program questions, and generate reports; an example workflow also sends reminders by email or tools such as Slack. Access follows each user's existing Security Journey permissions, and setup is through an admin connections page, with customer success support available. The integrations use MCP, are part of Aspen's Assist capability, and are included at no additional cost for Security Journey customers. The excerpt says learner support in the assistants and connections to more AI assistants are planned; it does not describe them as available now.

### Source excerpt

Secure code and AI workforce training provider Security Journey announced Nov. 12 that it has launched integrations with Claude and ChatGPT, built on the Model Context Protocol (MCP). Program owners can now check on learners, assign training and pull reports in plain language from inside the AI assistants they already use, instead of logging in... ... continue reading The post Security Journey brings training program management into Claude and ChatGPT appeared first on SD Times.

## MCP for Software Testing: How It Could Change AI-Powered QA

DevFeed: [MCP for Software Testing: How It Could Change AI-Powered QA](<https://devfeed.tech/articles/mcp-for-software-testing-how-it-could-change-ai-powered-qa-136320.md>)

Original publisher: [Read original article](<https://testgrid.io/blog/mcp-for-software-testing/>)

Author: Jayesh Karkar

Published: 2026-11-16T12:10:47Z

Content type: article

Language: en

Sources: [TestGrid](<https://devfeed.tech/sources/testgrid-blog.md>)

Topics: [Software Testing](<https://devfeed.tech/topics/software-testing.md>), [Test automation](<https://devfeed.tech/topics/test-automation.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [issue tracking](<https://devfeed.tech/topics/issue-tracking.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [can](<https://devfeed.tech/tags/can.md>), [change](<https://devfeed.tech/tags/change.md>), [code](<https://devfeed.tech/tags/code.md>), [context](<https://devfeed.tech/tags/context.md>), [defects](<https://devfeed.tech/tags/defects.md>), [dependencies](<https://devfeed.tech/tags/dependencies.md>), [environments](<https://devfeed.tech/tags/environments.md>), [failures](<https://devfeed.tech/tags/failures.md>), [help](<https://devfeed.tech/tags/help.md>), [large](<https://devfeed.tech/tags/large.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-for-software-testing](<https://devfeed.tech/tags/mcp-for-software-testing.md>), [model](<https://devfeed.tech/tags/model.md>), [model-context-protocol](<https://devfeed.tech/tags/model-context-protocol.md>), [model-context-protocol-mcp](<https://devfeed.tech/tags/model-context-protocol-mcp.md>), [qa](<https://devfeed.tech/tags/qa.md>), [software](<https://devfeed.tech/tags/software.md>), [software-testing](<https://devfeed.tech/tags/software-testing.md>), [testing](<https://devfeed.tech/tags/testing.md>)

### AI overview

MCP provides a common way for AI applications to discover and invoke tools, resources, and prompts exposed by servers, helping coordinate QA work across connected systems. It does not make existing tools or test suites available automatically: each system needs a compatible MCP server or adapter, and access depends on host permissions and authorization. Examples include generating test cases from code and requirements, browser automation, correlating failures with logs or code changes, and creating defect tickets. Adoption involves checking host and server compatibility, authentication and security, verifying tool access, and piloting workflows. Teams should validate AI interpretations against requirements and execution evidence, maintain integrations as protocol and tool definitions change, and handle outages carefully; retries are appropriate only for operations known to be safe or idempotent. Existing automation frameworks can work alongside MCP when an integration exposes the needed capabilities.

### Source excerpt

When a code change affects multiple modules, you have to determine which regression tests you need to rerun, check dependencies, investigate failures, and trace defects across different environments. And if your QA team is managing large suites, interconnecting context and tools across these activities can be operationally cumbersome. Model Context Protocol (MCP) can help by [...]

## Coding agent, meet ngrok: the ngrok MCP server is live

DevFeed: [Coding agent, meet ngrok: the ngrok MCP server is live](<https://devfeed.tech/articles/coding-agent-meet-ngrok-the-ngrok-mcp-server-is-live-136004.md>)

Original publisher: [Read original article](<https://ngrok.com/blog/mcp>)

Author: Joel Hans

Published: 2026-11-16T00:00:00Z

Content type: release

Language: en

Sources: [\[ngrok news\]](<https://devfeed.tech/sources/ngrok-news.md>)

Topics: [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [cursor](<https://devfeed.tech/topics/cursor.md>), [codex](<https://devfeed.tech/topics/codex.md>), [Claude Code](<https://devfeed.tech/topics/claude-code.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [2026](<https://devfeed.tech/tags/2026.md>), [3](<https://devfeed.tech/tags/3.md>), [404](<https://devfeed.tech/tags/404.md>), [about](<https://devfeed.tech/tags/about.md>), [account](<https://devfeed.tech/tags/account.md>), [agent](<https://devfeed.tech/tags/agent.md>), [ai](<https://devfeed.tech/tags/ai.md>), [code](<https://devfeed.tech/tags/code.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agent](<https://devfeed.tech/tags/coding-agent.md>), [free](<https://devfeed.tech/tags/free.md>), [internet](<https://devfeed.tech/tags/internet.md>), [live](<https://devfeed.tech/tags/live.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-server](<https://devfeed.tech/tags/mcp-server.md>), [needs](<https://devfeed.tech/tags/needs.md>), [ngrok](<https://devfeed.tech/tags/ngrok.md>), [plan](<https://devfeed.tech/tags/plan.md>), [product-updates](<https://devfeed.tech/tags/product-updates.md>), [server](<https://devfeed.tech/tags/server.md>), [traffic-policy](<https://devfeed.tech/tags/traffic-policy.md>), [works](<https://devfeed.tech/tags/works.md>)

### AI overview

ngrok's MCP server lets compatible coding agents inspect traffic and account resources, search current documentation, and create or update cloud endpoint Traffic Policy rules. In the gpu.compare example, the agent removed an OAuth gate, analyzed traffic, added scanner blocks and a per-IP rate limit, and configured a country cookie; it validated policy rules and checked their effects. The server provides tool descriptions, documentation search, and a policy validator to reduce configuration guesswork. Access is scope-based: users approve read or write permissions, and OAuth tokens are limited to the MCP server, expire after 15 minutes with refresh during work, and require renewed approval after a week. The server does not accept API keys or authtokens; a personal access token is an option for clients without OAuth. The MCP manages account-side configuration, while the ngrok CLI still creates tunnels. Availability and supported clients vary by the listed setup options; ChatGPT is stated not to connect yet.

### Source excerpt

The ngrok MCP server is live. Say what needs to happen between your code and the internet, and your coding agent turns it into ngrok. Works on every plan, even free.

## Qwen4-Max now available on AI Gateway

DevFeed: [Qwen4-Max now available on AI Gateway](<https://devfeed.tech/articles/qwen4-max-now-available-on-ai-gateway-134730.md>)

Original publisher: [Read original article](<https://vercel.com/changelog/qwen4-max-now-available-on-ai-gateway>)

Author: Zachary Chen

Published: 2026-11-15T20:00:00Z

Content type: release

Language: en

Sources: [Vercel News](<https://devfeed.tech/sources/vercel-news.md>)

Topics: [qwen](<https://devfeed.tech/topics/qwen.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [context window](<https://devfeed.tech/topics/context-window.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [LLM security](<https://devfeed.tech/topics/llm-security.md>), [Memory Safety](<https://devfeed.tech/topics/memory-safety.md>)

Tags: [3d](<https://devfeed.tech/tags/3d.md>), [access-control](<https://devfeed.tech/tags/access-control.md>), [agent-tasks](<https://devfeed.tech/tags/agent-tasks.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-gateway](<https://devfeed.tech/tags/ai-gateway.md>), [alibaba](<https://devfeed.tech/tags/alibaba.md>), [animation](<https://devfeed.tech/tags/animation.md>), [blender](<https://devfeed.tech/tags/blender.md>), [bring-your-own-key](<https://devfeed.tech/tags/bring-your-own-key.md>), [build](<https://devfeed.tech/tags/build.md>), [claude](<https://devfeed.tech/tags/claude.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [cli](<https://devfeed.tech/tags/cli.md>), [code](<https://devfeed.tech/tags/code.md>), [codex](<https://devfeed.tech/tags/codex.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [context](<https://devfeed.tech/tags/context.md>), [context-window](<https://devfeed.tech/tags/context-window.md>), [qwen4-max](<https://devfeed.tech/tags/qwen4-max.md>)

### AI overview

Alibaba's Qwen4-Max is available through AI Gateway as a reasoning model with native vision and a 1M-token context window. The supplied text says it can use visual feedback to refine Blender scenes and games built in Unity, Unreal Engine, or Godot; call MCP tools and process document and image collections for professional work; and deliver an early version of long-running tasks for review before making successive improvements. Alibaba is open-sourcing a Live Delivery plugin for this workflow. Compared with prior Qwen Max models, it is claimed to use fewer tokens for the same tasks and produce more secure code, with fewer injection, memory safety, and access control vulnerabilities. Use the model name `alibaba/qwen4-max`; setup for Claude Code, Codex, Cursor, and other agents uses the latest Vercel CLI. AI Gateway offers usage and cost tracking, retries, failover, and other routing features, with provider pricing, no markup, and no inference platform fee, including for BYOK requests.

### Source excerpt

Qwen4-Max from Alibaba is now available on AI Gateway. It is the first model in the Qwen4 series: a reasoning model with native vision and a 1M-token context window, built for coding, professional work, and long-running agent tasks. Qwen4-Max is better at working with professional tools. It can build 3D scenes in Blender and playable games in Unity, Unreal Engine, and Godot, running what it builds and using visual feedback to fix camera, animation, and lighting issues. For finance, legal, and healthcare work, it can call many MCP tools, work through large sets of documents and images, and produce spreadsheets, slide decks, and reports that cite their sources. For long-running tasks, the model delivers a usable first version early and then improves it in successive releases, so you can review progress and change direction before the work is finished. Alibaba is open-sourcing a Live Delivery plugin for this workflow. Compared with prior Qwen Max models, Qwen4-Max uses fewer tokens to complete the same tasks and writes more secure code by default, with fewer injection, memory safety, and access control vulnerabilities. Use alibaba/qwen4-max as the model name: To use it in Claude Code, Codex, Cursor, and more, install the latest Vercel CLI and run setup: Then select alibaba/qwen4-max in the agent. See the coding agents guide for details. Try Qwen4-Max in the model playground. AI Gateway provides a unified API for calling models, tracking usage and cost, and configuring retries, failover, and performance optimizations for higher-than-provider uptime. It includes built-in custom reporting, Zero Data Retention support, budgets for API keys, routing rules, and more. AI Gateway reflects provider pricing with no markup and does not charge a platform fee on inference, including on Bring Your Own Key (BYOK) requests. You can view all language models available on AI Gateway. Read more

## A Week of Symfony #1037 (November 9-15, 2026)

DevFeed: [A Week of Symfony #1037 (November 9-15, 2026)](<https://devfeed.tech/articles/a-week-of-symfony-1037-november-9-15-2026-134387.md>)

Original publisher: [Read original article](<https://symfony.com/blog/a-week-of-symfony-1037-november-9-15-2026>)

Author: Javier Eguiluz

Published: 2026-11-15T08:16:00Z

Content type: news

Language: en

Sources: [Symfony](<https://devfeed.tech/sources/symfony-blog.md>)

Topics: [Symfony](<https://devfeed.tech/topics/symfony.md>), [project release](<https://devfeed.tech/topics/project-release.md>), [Test automation](<https://devfeed.tech/topics/test-automation.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [Publish-subscribe pattern](<https://devfeed.tech/topics/pubsub.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [ai](<https://devfeed.tech/tags/ai.md>), [articles](<https://devfeed.tech/tags/articles.md>), [backend](<https://devfeed.tech/tags/backend.md>), [beta](<https://devfeed.tech/tags/beta.md>), [code](<https://devfeed.tech/tags/code.md>), [features](<https://devfeed.tech/tags/features.md>), [form](<https://devfeed.tech/tags/form.md>), [improvements](<https://devfeed.tech/tags/improvements.md>), [messenger](<https://devfeed.tech/tags/messenger.md>), [release](<https://devfeed.tech/tags/release.md>), [symfony](<https://devfeed.tech/tags/symfony.md>)

### AI overview

Symfony published the second beta of Symfony 8.2 for testing, alongside articles on its features, and announced the Symfony AI & MCP Hackathon drew around 50 teams. SymfonyLive Paris 2027 is scheduled for March 26-27 at Espace 42, with more SymfonyCon Warsaw 2026 talks announced. The project merged 197 pull requests and closed 38 issues this week. The listed changes include fixes across routing, security, locking, rate limiting, Messenger, HTTP clients, caching, and other components, as well as AI-related additions such as configurable MCP request body limits and improved agent tool handling. This overview reflects the supplied excerpt and selected change list.

### Source excerpt

This week, we published the second beta of Symfony 8.2 so you can keep testing it before its final release. Meanwhile, we kept publishing articles about Symfony 8.2 features, such as Messenger flows, AssetMapper improvements, form improvements, validator...

## How to classify MCP tools by risk (read, edit, delete)

DevFeed: [How to classify MCP tools by risk (read, edit, delete)](<https://devfeed.tech/articles/how-to-classify-mcp-tools-by-risk-read-edit-delete-133517.md>)

Original publisher: [Read original article](<https://nirmata.com/2026/11/16/classifying-mcp-tools-by-risk/>)

Author: Ritesh Patel

Published: 2026-11-14T03:15:26.237377Z

Content type: article

Language: en

Sources: [Nirmata](<https://devfeed.tech/sources/nirmata.md>)

Topics: [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [CRUD](<https://devfeed.tech/topics/crud.md>), [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-governance](<https://devfeed.tech/tags/ai-governance.md>), [class](<https://devfeed.tech/tags/class.md>), [delete](<https://devfeed.tech/tags/delete.md>), [how-to](<https://devfeed.tech/tags/how-to.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-tools](<https://devfeed.tech/tags/mcp-tools.md>), [risk](<https://devfeed.tech/tags/risk.md>), [security](<https://devfeed.tech/tags/security.md>), [tools](<https://devfeed.tech/tags/tools.md>)

### AI overview

Classify MCP tools by reviewed behavior: read tools change nothing, edit tools add or update state, and delete tools may destroy or overwrite it. `readOnlyHint` and `destructiveHint` help map tools to these classes, but annotations are optional, may be inaccurate, and must not drive client decisions on their own. With no hints, treat a tool as potentially destructive until checked. List the server's tools, record its hints, inspect documentation or code, and assign each tool the worst-case class. Track `openWorldHint` separately because a read tool may reach external entities. The article maps read tools to allow rules, edits to prompts, and deletes to deny by default, while emphasizing that client rules do not replace server-side access controls.

### Source excerpt

Sort MCP tools into read, edit and delete classes with the spec's readOnlyHint and destructiveHint. Set each class yourself, since hints are not guarantees.

## BYOH: Bring Your Own Harness with Cisco IQ, for partners

DevFeed: [BYOH: Bring Your Own Harness with Cisco IQ, for partners](<https://devfeed.tech/articles/byoh-bring-your-own-harness-with-cisco-iq-for-partners-132459.md>)

Original publisher: [Read original article](<https://blogs.cisco.com/customerexperience/byoh-bring-your-own-harness-with-cisco-iq-for-partners>)

Author: Carlos Pereira

Published: 2026-11-13T16:22:46Z

Content type: article

Language: en

Sources: [Cisco Blogs](<https://devfeed.tech/sources/cisco-blogs.md>)

Topics: [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [AI-Ready Data](<https://devfeed.tech/topics/ai-ready-data.md>), [Cisco](<https://devfeed.tech/topics/cisco.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>)

Tags: [2025](<https://devfeed.tech/tags/2025.md>), [2026](<https://devfeed.tech/tags/2026.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-buildout](<https://devfeed.tech/tags/ai-buildout.md>), [ai-models](<https://devfeed.tech/tags/ai-models.md>), [ai-native](<https://devfeed.tech/tags/ai-native.md>), [ai-risk](<https://devfeed.tech/tags/ai-risk.md>), [api](<https://devfeed.tech/tags/api.md>), [apis](<https://devfeed.tech/tags/apis.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [assets](<https://devfeed.tech/tags/assets.md>), [bring-your-own-harness](<https://devfeed.tech/tags/bring-your-own-harness.md>), [cisco](<https://devfeed.tech/tags/cisco.md>), [cisco-customer-experience](<https://devfeed.tech/tags/cisco-customer-experience.md>), [cisco-iq](<https://devfeed.tech/tags/cisco-iq.md>), [cisco-partners](<https://devfeed.tech/tags/cisco-partners.md>), [context](<https://devfeed.tech/tags/context.md>), [harness](<https://devfeed.tech/tags/harness.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [partners](<https://devfeed.tech/tags/partners.md>), [security](<https://devfeed.tech/tags/security.md>)

### AI overview

The article presents Cisco IQ as a trusted customer-context source for partners using their preferred AI harnesses. Through Cisco IQ APIs and MCP, harnesses can query information grounded in a customer's environment, subject to the partner's granted access; read and change permissions are separate. In an example using Claude Code, Codex, and Cursor, each reported around 59% of assets lacked support coverage, while their reasoning and observations differed. The article says MCP's open standard avoids locking partners to one AI vendor. It recommends discovering which harnesses teams already use, starting with read-only access before enabling changes, and treating trusted context as an asset. It identifies overreach and under-grounding as risks, and says partners and customers should decide together which model provider receives customer context and under what terms.

### Source excerpt

With bringing your own harness (BYOH) trusted context has become part of the security perimeter. Here's how Cisco IQ fits in.

## Take Claude Code Above Average: Let Antigravity Review It

DevFeed: [Take Claude Code Above Average: Let Antigravity Review It](<https://devfeed.tech/articles/take-claude-code-above-average-let-antigravity-review-it-132382.md>)

Original publisher: [Read original article](<https://medium.com/google-developer-experts/take-claude-code-above-average-let-antigravity-review-it-a945412e7dcf?source=rss----a67bd6fa7d58---4>)

Author: Daniel Gwerzman

Published: 2026-11-13T14:21:02Z

Content type: opinion

Language: en

Sources: [Google Developer Experts](<https://devfeed.tech/sources/google-developer-experts-medium.md>)

Topics: [Code review](<https://devfeed.tech/topics/code-review.md>), [ai code review](<https://devfeed.tech/topics/ai-code-review.md>), [Claude Code](<https://devfeed.tech/topics/claude-code.md>), [google-antigravity](<https://devfeed.tech/topics/google-antigravity.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [Claude](<https://devfeed.tech/topics/claude.md>), [codex](<https://devfeed.tech/topics/codex.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [antigravity](<https://devfeed.tech/tags/antigravity.md>), [bridge](<https://devfeed.tech/tags/bridge.md>), [claude](<https://devfeed.tech/tags/claude.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [code](<https://devfeed.tech/tags/code.md>), [code-review](<https://devfeed.tech/tags/code-review.md>), [design](<https://devfeed.tech/tags/design.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [gemini-4-argon](<https://devfeed.tech/tags/gemini-4-argon.md>), [google-antigravity](<https://devfeed.tech/tags/google-antigravity.md>), [google-developer-expert](<https://devfeed.tech/tags/google-developer-expert.md>), [november-2026](<https://devfeed.tech/tags/november-2026.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [process](<https://devfeed.tech/tags/process.md>), [review](<https://devfeed.tech/tags/review.md>), [security](<https://devfeed.tech/tags/security.md>), [stage](<https://devfeed.tech/tags/stage.md>), [tests](<https://devfeed.tech/tags/tests.md>)

### AI overview

The author's central finding is that independent review works best as a dialogue between different AI tools and models, with a human resolving disagreements. In the described setup, Claude Code leads while Antigravity with Gemini 4 Argon reviews; a second review caught concurrent-session risk and missing backups, and later discussion exposed a proposed fix that would fail from a subfolder. The four-stage loop briefs an outside reviewer, requests structured findings and evidence for assumptions, lets the author challenge findings once, then sends changes back for verification and escalates unresolved questions. The author's agy-bridge MCP server connects Claude Code to Antigravity's headless `agy` CLI, preserving review conversations and backing up and comparing project state each turn. Reviewers can read files and run tests but are instructed to write only in scratch space; shell commands cannot be similarly restricted, and project `.agents/` skills may steer review, so the author advises checking them in untrusted repositories. The method is intended for consequential work; reviews take minutes and are unnecessary for throwaway tasks.

### Source excerpt

In November 2026, I spent a day building a small tool with Claude Code: a bridge that lets a second AI agent review Claude's work, run the tests in my project, and argue about what it finds. Partway through the day, the design was revised, the code was written, the tests passed, and Claude was satisfied with its work. Then I gave the same design and code to a different model, running in a different tool. It came back with two findings marked high. Two terminal windows could drive the same review at the same time, each starting its own process behind the other's back. And the reviewer could run shell commands in my project, but nothing backed up my uncommitted work before its turn. One stray git reset --hard would have wiped it. Both were real. Both were cheap to fix. Neither was the kind of thing the author finds by "double-checking". That was the first fix of the day. There were two more. One more thing about that reviewer. It wasn't a second Claude session. It was Antigravity, Google's agentic development platform, running a Gemini model. Today, every review I run goes to Gemini 4 Argon. That pairing is what this article is about: why Claude Code does better work when Antigravity reviews it, the four-stage review I run between them, and the open-source bridge I built so Claude Code can call Antigravity on its own. Why the Author Can't Be the Reviewer Start with what these models are. They learned from everything people have put online: the brilliant answer and the one that ships a security hole, the masterpiece and the first draft. They learned the distribution. So by default, they give you the middle of it. Ask for a function and you get the average function: it works on the happy path and carries the usual number of subtle problems. That isn't a flaw. It's the mathematics of the system. Average is the floor, not the ceiling. The same model has seen far better work than the average. But getting it there takes a process, not a cleverer prompt. The second thing to

## Your pinned MCP server may not be as pinned as you think

DevFeed: [Your pinned MCP server may not be as pinned as you think](<https://devfeed.tech/articles/your-pinned-mcp-server-may-not-be-as-pinned-as-you-think-131237.md>)

Original publisher: [Read original article](<https://www.helpnetsecurity.com/2026/11/13/corelight-mcp-server-supply-chain/>)

Author: Mirko Zorz

Published: 2026-11-13T05:30:22Z

Content type: news

Language: en

Sources: [Help Net Security](<https://devfeed.tech/sources/help-net-security.md>)

Topics: [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [code security](<https://devfeed.tech/topics/code-security.md>), [Python](<https://devfeed.tech/topics/python.md>), [versioning](<https://devfeed.tech/topics/versioning.md>)

Tags: [ai-coding-assistants](<https://devfeed.tech/tags/ai-coding-assistants.md>), [automation](<https://devfeed.tech/tags/automation.md>), [corelight](<https://devfeed.tech/tags/corelight.md>), [credentials](<https://devfeed.tech/tags/credentials.md>), [cybersecurity](<https://devfeed.tech/tags/cybersecurity.md>), [dependencies](<https://devfeed.tech/tags/dependencies.md>), [don-t-miss](<https://devfeed.tech/tags/don-t-miss.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-server](<https://devfeed.tech/tags/mcp-server.md>), [news](<https://devfeed.tech/tags/news.md>), [pinned](<https://devfeed.tech/tags/pinned.md>), [security](<https://devfeed.tech/tags/security.md>), [supply-chain-attacks](<https://devfeed.tech/tags/supply-chain-attacks.md>)

### AI overview

A fixed MCP server release did not keep its dependencies fixed: over five weeks, an administrator's workstation built 15 dependency sets without configuration changes, and 23 of 69 packages changed versions. The article attributes this to `uvx` installing declared dependencies afresh while skipping a community server's hash-pinned lock file; pinning the top-level release alone therefore did not pin its dependency tree. Corelight found that 23 of 35 reviewed servers had recommended installs that could pull from a public registry on launch, install or image build. In a two-week trial, five servers installed once from hash-pinned lock files and launched directly retained matching package lists across 457 starts. The tradeoff is that frozen dependencies can age: one frozen server had 40 packages older than what a fresh install would have selected. Lagermann recommends following maintainers' update guidance and describes nightly vulnerability reviews for servers he maintains.

### Source excerpt

An MCP server configured to run one fixed release on an administrator's workstation ran with 15 different sets of dependencies over five weeks, with its configuration untouched. James Lagermann, Director of Technical Alliances at Corelight, found the drift in his company's lab. Twenty-three of the server's 69 packages changed versions, including the cryptography library, the JSON Web Token library and the MCP software development kit itself. MCP (Model Context Protocol) servers let AI coding assistants ... More -> The post Your pinned MCP server may not be as pinned as you think appeared first on Help Net Security.

## Turn Jira tickets into GitLab merge requests with Duo Agent Platform

DevFeed: [Turn Jira tickets into GitLab merge requests with Duo Agent Platform](<https://devfeed.tech/articles/turn-jira-tickets-into-gitlab-merge-requests-with-duo-agent-platform-131497.md>)

Original publisher: [Read original article](<https://about.gitlab.com/blog/jira-ticket-to-gitlab-mr-with-duo-agent-platform/>)

Author: Itzik Gan Baruch

Published: 2026-11-12T00:00:00Z

Content type: tutorial

Language: en

Sources: [GitLab](<https://devfeed.tech/sources/gitlab.md>)

Topics: [GitLab](<https://devfeed.tech/topics/gitlab.md>), [jira](<https://devfeed.tech/topics/jira.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [Pull Request](<https://devfeed.tech/topics/pull-request.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [admin](<https://devfeed.tech/tags/admin.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-catalog](<https://devfeed.tech/tags/ai-catalog.md>), [also](<https://devfeed.tech/tags/also.md>), [api](<https://devfeed.tech/tags/api.md>), [atlassian](<https://devfeed.tech/tags/atlassian.md>), [beta](<https://devfeed.tech/tags/beta.md>), [can](<https://devfeed.tech/tags/can.md>), [catalog](<https://devfeed.tech/tags/catalog.md>), [change](<https://devfeed.tech/tags/change.md>), [duo](<https://devfeed.tech/tags/duo.md>), [gitlab](<https://devfeed.tech/tags/gitlab.md>), [jira](<https://devfeed.tech/tags/jira.md>), [merge](<https://devfeed.tech/tags/merge.md>), [platform](<https://devfeed.tech/tags/platform.md>), [product](<https://devfeed.tech/tags/product.md>), [tickets](<https://devfeed.tech/tags/tickets.md>), [tutorial](<https://devfeed.tech/tags/tutorial.md>), [with](<https://devfeed.tech/tags/with.md>)

### AI overview

The tutorial connects Jira and GitLab so a custom Jira agent can search, summarize, and comment on tickets, then hand implementation to Developer Flow with `/goal`, which creates a draft merge request. GitLab's Jira integration links MRs that mention a ticket key back to the Jira ticket and can post comments. Setup requires GitLab Premium or Ultimate, Duo Agent Platform and Developer Flow, a Jira Cloud project, and an Atlassian API token for the standard integration; Atlassian organization admin access is needed to authorize GitLab as an MCP client. Each user connects their own Atlassian account, so Jira access follows their permissions and comments appear under their name. Developer Flow runs as a CI pipeline with Git and GitLab API write access after the goal is sent; review the draft MR, as nothing merges automatically. The Atlassian MCP server is experimental and not production-ready, and each goal consumes GitLab Credits.

### Source excerpt

Many teams plan in Jira and ship in GitLab. Even if you already use GitLab Duo Agent Platform, you still open Jira to read each ticket, copy it into an agent for implementation, and report the merge request back. This tutorial shows you how to remove that handoff. You add a Jira agent to Duo Agentic Chat. The agent knows your project and repository, and you use it to find, create, and update tickets. Then you hand a ticket to Developer Flow with /goal, and Developer Flow creates the merge request. Watch the short demo first: A custom Jira agent finds the urgent tickets, one goes to Developer Flow, and the merge request shows up on the Jira ticket. Prerequisites GitLab Premium or Ultimate on GitLab.com, with GitLab Duo Agent Platform turned on for your top-level group and Developer Flow enabled for the project. For the Atlassian MCP server in the AI Catalog: experiment and beta features turned on for the top-level group (Settings > GitLab Duo > Change configuration). Maintainer or Owner role on the project for the one-time settings in Steps 1 and 2; Developer or above on a project to use the agent there afterwards. A Jira Cloud project you can read and comment on, and Atlassian admin access once, to trust GitLab as a Model Context Protocol (MCP) client. Jira names a ticket with a ticket key: the project key plus a number. Our examples use project key KAN, so an example ticket key looks like KAN-123. The examples also use GitLab project my-group/my-project. A Jira API token from id.atlassian.com, used only by GitLab's Jira integration in Step 1. Note: The Atlassian MCP server in the AI Catalog is an experiment and isn't ready for production use. Try it on a test project first. Tutorial Here are the steps to follow to turn Jira tickets into GitLab merge requests. 1. Connect GitLab to Jira This is GitLab's standard Jira integration. It makes any MR that mentions a ticket, for example KAN-123, show up on that Jira ticket, and posts a comment there. Five minutes, done onc

## Genie One: The AI cowork home for everyone

DevFeed: [Genie One: The AI cowork home for everyone](<https://devfeed.tech/articles/genie-one-the-ai-cowork-home-for-everyone-129973.md>)

Original publisher: [Read original article](<https://www.databricks.com/blog/genie-one-ai-cowork-home-everyone>)

Author: Sydney Sundell; Elise Georis

Published: 2026-11-11T15:00:00Z

Content type: release

Language: en

Sources: [Databricks](<https://devfeed.tech/sources/databricks.md>)

Topics: [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [scheduling](<https://devfeed.tech/topics/scheduling.md>), [Tool](<https://devfeed.tech/topics/tool.md>), [tracing](<https://devfeed.tech/topics/tracing.md>)

Tags: [access-controls](<https://devfeed.tech/tags/access-controls.md>), [accountability](<https://devfeed.tech/tags/accountability.md>), [act](<https://devfeed.tech/tags/act.md>), [acting](<https://devfeed.tech/tags/acting.md>), [actions](<https://devfeed.tech/tags/actions.md>), [admin](<https://devfeed.tech/tags/admin.md>), [adoption](<https://devfeed.tech/tags/adoption.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-assistants](<https://devfeed.tech/tags/ai-assistants.md>), [ai-coworker](<https://devfeed.tech/tags/ai-coworker.md>), [ai-search](<https://devfeed.tech/tags/ai-search.md>), [ai-tools](<https://devfeed.tech/tags/ai-tools.md>), [all](<https://devfeed.tech/tags/all.md>), [cowork](<https://devfeed.tech/tags/cowork.md>), [enterprise](<https://devfeed.tech/tags/enterprise.md>), [genie](<https://devfeed.tech/tags/genie.md>), [home](<https://devfeed.tech/tags/home.md>), [platform](<https://devfeed.tech/tags/platform.md>), [product](<https://devfeed.tech/tags/product.md>), [work](<https://devfeed.tech/tags/work.md>)

### AI overview

Genie One combines enterprise search, governed data analysis, artifact creation, and cross-tool actions in one conversational workspace. Genie Search indexes knowledge from sources such as Google Drive, Jira, Confluence, Workday, and SharePoint, and can also use more than 100 MCP connectors for live lookups and actions; Genie chooses among search, MCP, and SQL based on the query. Users can create and refine decks, dashboards, and interactive pages with source visibility, while scheduled tasks keep artifact data current and skills apply shared workflows or brand standards. MCP connections let Genie update tools such as Jira and Asana, with users able to require approval for actions and admins controlling allowed connections. Unity Gateway provides action traces and policy controls, while Unity Catalog permissions and connected-source access controls limit what each user can access. Genie is also available through native apps and plugins for ChatGPT, Claude, Microsoft Copilot Cowork, and Grok Bot, with governed context and permissions.

### Source excerpt

In today's enterprise, work rarely stays in one application. A single task can involve...

## MCP and A2A alone won't scale AI agents

DevFeed: [MCP and A2A alone won't scale AI agents](<https://devfeed.tech/articles/mcp-and-a2a-alone-won-t-scale-ai-agents-129003.md>)

Original publisher: [Read original article](<https://www.infoworld.com/article/4246624/mcp-and-a2a-alone-wont-scale-ai-agents.html>)

Author: Isaac Sacolick

Published: 2026-11-11T09:00:00Z

Content type: article

Language: en

Sources: [InfoWorld](<https://devfeed.tech/sources/infoworld.md>)

Topics: [A2A protocol](<https://devfeed.tech/topics/a2a-protocol.md>), [Multi Agent Systems](<https://devfeed.tech/topics/multi-agent-systems.md>), [agent to agent communication](<https://devfeed.tech/topics/agent-to-agent-communication.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [agent orchestration](<https://devfeed.tech/topics/agent-orchestration.md>)

Tags: [a2a](<https://devfeed.tech/tags/a2a.md>), [agent-orchestration](<https://devfeed.tech/tags/agent-orchestration.md>), [agent2agent](<https://devfeed.tech/tags/agent2agent.md>), [agent2agent-protocol](<https://devfeed.tech/tags/agent2agent-protocol.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [ai-agent-orchestration](<https://devfeed.tech/tags/ai-agent-orchestration.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [artificial-intelligence-data-and-information-security-data-management-development-tools-devops](<https://devfeed.tech/tags/artificial-intelligence-data-and-information-security-data-management-development-tools-devops.md>), [claude](<https://devfeed.tech/tags/claude.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [clients](<https://devfeed.tech/tags/clients.md>), [code](<https://devfeed.tech/tags/code.md>), [coding](<https://devfeed.tech/tags/coding.md>), [collibra](<https://devfeed.tech/tags/collibra.md>), [connect](<https://devfeed.tech/tags/connect.md>), [connectors](<https://devfeed.tech/tags/connectors.md>), [context](<https://devfeed.tech/tags/context.md>), [data-and-information-security](<https://devfeed.tech/tags/data-and-information-security.md>), [data-management](<https://devfeed.tech/tags/data-management.md>), [development-tools](<https://devfeed.tech/tags/development-tools.md>), [devops](<https://devfeed.tech/tags/devops.md>), [generative-ai](<https://devfeed.tech/tags/generative-ai.md>), [language-en](<https://devfeed.tech/tags/language-en.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [scale](<https://devfeed.tech/tags/scale.md>), [security](<https://devfeed.tech/tags/security.md>), [software-development](<https://devfeed.tech/tags/software-development.md>), [storytype-analysis](<https://devfeed.tech/tags/storytype-analysis.md>)

### AI overview

MCP connects an agent to tools, data, and APIs; A2A lets independently built agents discover one another, negotiate capabilities, delegate tasks, and hand off work. The text argues that these protocols enable connections and collaboration but do not by themselves scale enterprise deployments. Organizations must separately define agent identities, permissions, and credential controls: MCP binds tokens to a server audience, while A2A stops at AUTH_REQUIRED and leaves credential acquisition to the client. Scaling also requires a governed, high quality shared context layer; redesigning business processes for agent collaboration; and observability and agentic operations for diagnosis, security, and audits. More endpoints and agent messages can move work faster without improving trust or shared learning, and may increase bad responses if implemented in silos. The article cites forecasts of more than 150,000 agents at the average Fortune 500 enterprise within two years and reports that 76% encountered critical pilot to production roadblocks in the prior 12 months.

### Source excerpt

Two protocols define how AI agents connect to internal resources and collaborate with other AI agents. The Model Context Protocol (MCP) is well known because it's promoted as a general protocol for connecting AI agents to resources and other AI agents. But that's not completely accurate. It's really the Agent2Agent protocol (A2A) that's enabling AI agent orchestration. The confusion stems from how quickly platforms have adopted these protocols and marketed them as features. Users building their own AI agents with Claude, OpenAI, Microsoft Copilot, and other models are already being exposed to customizing skills, connectors, and other plug-ins that use MCP and A2A. Vibe coding and AI app-building tools like Bolt.new, Lovable, Pave, Replit, and v0 can connect to MCP servers and include built-in connectors. AI coding editors like Claude Code, Cursor, Devin Desktop, Gemini CLI, GitHub Copilot, and Kiro have MCP clients. These protocols will shape how enterprises scale and manage AI agents. Gartner predicts that the average global Fortune 500 enterprise will run more than 150,000 AI agents over the next two years. But according to Collibra's 2026 Hallucination Tax Report, 76% have hit critical roadblocks while attempting to move from pilot to full production during the past 12 months. Reviewing the two protocols and which services are your responsibility to layer on top of them is key to scaling enterprise-ready AI agents. Differences between MCP and A2A Scott Gerlach, cofounder and chief security officer at StackHawk, offers this analogy to help technologists better understand MCP and A2A's functions. "If you think about doing some bathroom remodeling and you call a plumber to move a sink, he'd use tools in his MCP toolbox to complete his work. But when it's time to fix the drywall, repaint, and put the vanity back together, the plumber agent would use A2A to call other contractor agents to complete that work," Gerlach says. Another approach to explaining MCP and A2A co

## How to Build an MCP Server That Checks npm Packages Before AI Installs Them

DevFeed: [How to Build an MCP Server That Checks npm Packages Before AI Installs Them](<https://devfeed.tech/articles/how-to-build-an-mcp-server-that-checks-npm-packages-before-ai-installs-them-128451.md>)

Original publisher: [Read original article](<https://hackernoon.com/how-to-build-an-mcp-server-that-checks-npm-packages-before-ai-installs-them?source=rss>)

Author: Tarun Singh

Published: 2026-11-11T07:39:09Z

Content type: tutorial

Language: en

Sources: [HackerNoon](<https://devfeed.tech/sources/hackernoon.md>)

Topics: [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [npm packages](<https://devfeed.tech/topics/npm-packages.md>), [Cloudflare Workers](<https://devfeed.tech/topics/cloudflare-workers.md>), [npm security](<https://devfeed.tech/topics/npm-security.md>), [npm](<https://devfeed.tech/topics/npm.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-coding-assistants](<https://devfeed.tech/tags/ai-coding-assistants.md>), [api](<https://devfeed.tech/tags/api.md>), [api-keys](<https://devfeed.tech/tags/api-keys.md>), [audit](<https://devfeed.tech/tags/audit.md>), [build](<https://devfeed.tech/tags/build.md>), [cloudflare](<https://devfeed.tech/tags/cloudflare.md>), [cloudflare-workers](<https://devfeed.tech/tags/cloudflare-workers.md>), [deprecation](<https://devfeed.tech/tags/deprecation.md>), [how-to](<https://devfeed.tech/tags/how-to.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-server](<https://devfeed.tech/tags/mcp-server.md>), [model-context-protocol](<https://devfeed.tech/tags/model-context-protocol.md>), [npm](<https://devfeed.tech/tags/npm.md>), [npm-packages](<https://devfeed.tech/tags/npm-packages.md>), [packages](<https://devfeed.tech/tags/packages.md>), [remote](<https://devfeed.tech/tags/remote.md>), [remote-mcp-server](<https://devfeed.tech/tags/remote-mcp-server.md>), [server](<https://devfeed.tech/tags/server.md>), [software-supply-chain](<https://devfeed.tech/tags/software-supply-chain.md>), [typescript](<https://devfeed.tech/tags/typescript.md>), [vulnerabilities](<https://devfeed.tech/tags/vulnerabilities.md>), [workers](<https://devfeed.tech/tags/workers.md>)

### AI overview

The article builds pkgscout, a stateless MCP server on Cloudflare Workers that checks npm packages before an AI suggests installing them. Its `inspect_package` tool reports package metadata and deprecation, and generates bounded typo variants when a name is missing; npm search can bury deprecated packages, while typo matches may themselves be misleading or malicious, so suggestions include warnings. `check_vulnerabilities` queries npm's advisory database but requires exact versions, rejecting ranges. To stay within Workers' 10 ms CPU limit, the server fetches compact `/latest` manifests instead of large package documents and caches registry calls for five minutes. The excerpt also explains bearer-token authentication, local testing with the SDK's modern protocol negotiation, and deployment. It describes a shared token as suitable for personal use and points to OAuth for team sharing; PyPI support and checks for newly created packages remain proposed additions.

### Source excerpt

Build a remote MCP server on Cloudflare Workers that checks npm packages for typos, deprecation and known vulnerabilities before your AI installs them.

## MCP vs CLI vs Skills: What Each Costs in Context

DevFeed: [MCP vs CLI vs Skills: What Each Costs in Context](<https://devfeed.tech/articles/mcp-vs-cli-vs-skills-what-each-costs-in-context-129100.md>)

Original publisher: [Read original article](<https://openrouter.ai/blog/insights/mcp-vs-cli-vs-skills/>)

Author: OpenRouter

Published: 2026-11-11T00:00:00Z

Content type: comparison

Language: en

Sources: [OpenRouter](<https://devfeed.tech/sources/openrouter.md>)

Topics: [context window](<https://devfeed.tech/topics/context-window.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [Agent Skills](<https://devfeed.tech/topics/agent-skills.md>), [Command-line interface](<https://devfeed.tech/topics/cli.md>), [Model Routing](<https://devfeed.tech/topics/model-routing.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agent-sdk](<https://devfeed.tech/tags/agent-sdk.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [ai-agent-context](<https://devfeed.tech/tags/ai-agent-context.md>), [anthropic](<https://devfeed.tech/tags/anthropic.md>), [anthropic-claude](<https://devfeed.tech/tags/anthropic-claude.md>), [anthropic-claude-haiku-5-5](<https://devfeed.tech/tags/anthropic-claude-haiku-5-5.md>), [api](<https://devfeed.tech/tags/api.md>), [caching](<https://devfeed.tech/tags/caching.md>), [cli](<https://devfeed.tech/tags/cli.md>), [context](<https://devfeed.tech/tags/context.md>), [context-window](<https://devfeed.tech/tags/context-window.md>), [costs](<https://devfeed.tech/tags/costs.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-server](<https://devfeed.tech/tags/mcp-server.md>), [model-context-protocol](<https://devfeed.tech/tags/model-context-protocol.md>), [skills](<https://devfeed.tech/tags/skills.md>), [tokens](<https://devfeed.tech/tags/tokens.md>), [tool](<https://devfeed.tech/tags/tool.md>)

### AI overview

In one GitHub task, the CLI used far fewer input tokens than the GitHub MCP server: on Sonnet 5.5, 1,421 versus 59,417 with 46 default tools; the skill used 3,315. Tool definitions and returned data both add cost: a full release object added about 20,400 tokens, while requesting only needed fields reduced issue results. Scoping MCP to two tools cut its first request, and a 4,000-character result cap reduced total task tokens, though that cap relied on fields appearing near the result's top. OpenRouter Tool Search reduced up-front definition costs to under 600 tokens, but whole-task results varied by model: it raised Haiku's cost and produced search retries or tool-call errors in some runs. The article recommends using a CLI when available, or scoped MCP with small results, and measuring on the intended model and provider. These are results from one server and one task.

### Source excerpt

Every tool you give an agent sends its definition to the model on every request, and you pay for it as input tokens. A Model Context Protocol (MCP) server can add tens of thousands of tokens to the context window. The usual alternatives are a command-line tool (CLI) or a skill.

## Every public API should have a feedback endpoint for agents

DevFeed: [Every public API should have a feedback endpoint for agents](<https://devfeed.tech/articles/every-public-api-should-have-a-feedback-endpoint-for-agents-129376.md>)

Original publisher: [Read original article](<https://freek.dev/3222-every-public-api-should-have-a-feedback-endpoint-for-agents>)

Author: Freek Van der Herten (freek@spatie.be)

Published: 2026-11-10T21:19:45Z

Content type: article

Language: en

Sources: [Freek](<https://devfeed.tech/sources/freek-dev-all-blogposts.md>)

Topics: [API](<https://devfeed.tech/topics/api.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>), [human in the loop agent](<https://devfeed.tech/topics/human-in-the-loop-agent.md>)

Tags: [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [api](<https://devfeed.tech/tags/api.md>), [feedback](<https://devfeed.tech/tags/feedback.md>), [mcp](<https://devfeed.tech/tags/mcp.md>)

### AI overview

Inspired by a proposal for agent-facing API feedback, the author added a tool to the MCP servers for Oh Dear, There There, and Mailcoach. When an agent cannot complete a task because MCP or API functionality is missing, it can send feedback directly to the provider. The excerpt describes reporting missing functionality; it does not specify how feedback is processed or whether it leads to changes.

### Source excerpt

I like this idea very much. Inspired by this tweet, I added a tool to the MCP servers for Oh Dear, There There, and Mailcoach that lets agents report missing MCP or API functionality. If an agent can't complete a task because something is missing, it can send that feedback directly to us. J Jack Sharkey @jsharkey 𝕏 Every public API should have a feedback endpoint for agents When my agent struggles to use your API, it should be able to submit structured feedback directly Your agent should consume that feedback, fix the API + docs, ship an update, and notify my agent Remove the human from the loop and let the agents work! Worth sharing Read on X ↗

## What's new in Runpod for Agents: November 2026

DevFeed: [What's new in Runpod for Agents: November 2026](<https://devfeed.tech/articles/what-s-new-in-runpod-for-agents-november-2026-127094.md>)

Original publisher: [Read original article](<https://www.runpod.io/blog/runpod-agent-skills-mcp-server>)

Author: Bohdan Yurov

Published: 2026-11-10T15:32:46Z

Content type: article

Language: en

Sources: [Runpod Blog.](<https://devfeed.tech/sources/runpod-blog.md>)

Topics: [Agent Skills](<https://devfeed.tech/topics/agent-skills.md>), [hardened ai agent skills](<https://devfeed.tech/topics/hardened-ai-agent-skills.md>), [context-architecture](<https://devfeed.tech/topics/context-architecture.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>), [Test-driven development](<https://devfeed.tech/topics/tdd.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [agents](<https://devfeed.tech/tags/agents.md>), [api](<https://devfeed.tech/tags/api.md>), [ci](<https://devfeed.tech/tags/ci.md>), [cli](<https://devfeed.tech/tags/cli.md>), [cold-start](<https://devfeed.tech/tags/cold-start.md>), [comfyui](<https://devfeed.tech/tags/comfyui.md>), [credentials](<https://devfeed.tech/tags/credentials.md>), [data-center](<https://devfeed.tech/tags/data-center.md>), [data-centers](<https://devfeed.tech/tags/data-centers.md>), [dc](<https://devfeed.tech/tags/dc.md>), [deploy](<https://devfeed.tech/tags/deploy.md>), [evals](<https://devfeed.tech/tags/evals.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-server](<https://devfeed.tech/tags/mcp-server.md>), [november-2026](<https://devfeed.tech/tags/november-2026.md>), [rules](<https://devfeed.tech/tags/rules.md>), [skills](<https://devfeed.tech/tags/skills.md>), [tests](<https://devfeed.tech/tags/tests.md>), [what-s-new](<https://devfeed.tech/tags/what-s-new.md>)

### AI overview

Runpod's agent setup pairs a hosted MCP server, which exposes typed infrastructure operations, with Markdown skills that guide tool choice and procedures. The article's central finding is that skills go stale unless their claims are sourced and tested. Runpod describes a concept graph whose rules cite documentation, API references or probes, plus build checks, more than 80 skill eval cases, and 52 live-account user stories run with and without skills. Live recipe runs and agent reports feed fixes back into the skills, rules and tests. One example rule delays creating a network volume until stock selection determines the Pod's data center, avoiding repeated volume deletion when trying another location. The article says this system helps agents diagnose issues from worker logs and follow correct volume and deployment procedures; it does not provide measured performance results.

### Source excerpt

Skills tell an agent what to do and the MCP server does it. See how sourced rules, evals and live-account tests keep Runpod's agent skills accurate.

[Next page](<https://devfeed.tech/topics/msp-mcp.md?cursor=WyIyMDI2LTExLTEwVDE1OjMyOjQ2KzAwOjAwIiwgIjFjODVhODQ5LTUyZDYtNDQ0Zi1iODQ5LThhZDY4NzgwZjUwYiJd>)