MCP's context problem, measured: tool definitions vs tool results. At Prague Crawl 2026, Apify CEO Jan Čurn gave a talk called MCP doesn't suck, your agent does. Critics of MCP, the Model Context Protocol, say it fills an agent's context window. His answer was that the protocol doesn't decide what goes into the context. The agent does, and "it's a mistake of the implementations if they put everything into the context". He named 2 costs. Tool definitions fill the context before the first question, and every tool result is added after it. I measured both on real MCP servers, then checked what production tools already do about them. Cost 1: tool definitions#. An agent that connects to an MCP server usually puts every tool's name, description and input schema into its context. I connected to 6 popular servers that run without an API key, listed their tools, and counted the tokens of those definitions. I counted with OpenAI's o200k_base tokenizer, so other models will differ a little. I ran it on October 12, 2026. | MCP server | Tools | Tokens | | Chrome DevTools MCP 1.10.1 | 30 | 6,527 | | Playwright MCP 0.0.83 | 25 | 4,450 | | Filesystem 1.0.0 | 14 | 1,957 | | Git 2026.10.10 | 12 | 1,426 | | Everything (reference server) 1.0.0 | 13 | 1,305 | | Fetch 2026.10.10 | 1 | 285 | | All 6 | 95 | 15,950 | 95 tools took 15,950 tokens, about 8% of a 200K-token context window. These are small servers, and larger ones cost far more. That's where the talk's warning that definitions can take "one third" of the context comes from. Anthropic's own example lists 5 servers, including GitHub's at 35 tools, that take about 55K tokens before the conversation starts. Anthropic has seen tool definitions take 134K tokens before optimization. This script measures it. It connects the way an agent does, over stdio. Cost 2: tool results#. Each tool call adds its result to the context too, and for web data tools the result is a page. I loaded 3 real pages with 2 tools and counted the tokens of what came back. | Page | Playwright page snapshot | Fetch, whole page | Fetch, default reply | | Wikipedia, "Web scraping" 31,565 | 11,468 | 1,067 | | Hacker News front page | 13,579 | 3,494 | 1,848 | | Allbirds product page | 10,824 | 1,217 | 1,217 | One snapshot of the Wikipedia article took 31,565 tokens, about twice as much as all 95 tool definitions together. An agent that reads a few pages fills its context with results, not definitions. What production tools already do#. The tools in this test already handle part of the problem. * Playwright MCP keeps the snapshot out of the context. Its browser_navigate tool saves the page snapshot to a .yml file and returns a short reply with a link to it. That reply took 78 to 177 tokens on these pages. The agent reads the snapshot only when it needs it. * The Fetch server returns 5,000 characters by default. That's the "default reply" column. The agent can ask for the next part, or raise the limit up to 999,999 characters. The talk named 3 more ways teams keep the context small. * Tool search. Instead of loading every tool, the agent gets one tool that searches the others and loads only the matches. Anthropic released this as the Tool Search Tool, and reported an 85% reduction in token usage in its own example. * Code mode. The agent writes code that calls the tools, and only the final result goes back to the model. Cloudflare's Code Mode works this way, because, in Cloudflare's words, "LLMs are better at writing code to call MCP, than at calling MCP directly". * A command-line client for MCP. Apify built mcpc, which turns any MCP server into shell commands an agent can run. It has a grep command for tool search across servers. I tried mcpc 0.7.0 with the 2 browser servers. A search for "screenshot" returned 4 matching tools, and loading the full details of one of them cost 454 tokens in total. Loading all 55 tools from both servers costs 10,977. What this means for a web data agent#. * Count both costs. Tool definitions are a fixed cost per session. Page results grow with every page the agent reads, and they were the larger cost in this test. * Keep pages out of the context. Save snapshots and pages to files, return a short reply, and extract only the fields you need, the way Playwright MCP and code mode do. * Load tools on demand when you connect many servers. With a few small servers, the definitions cost a few thousand tokens. With large servers like GitHub's, tool search saves much more. Limits of this test#. * One tokenizer. o200k_base is OpenAI's tokenizer. Claude and other models count a little differently. * Servers that run without a key. Servers that need an account, like Apify's or GitHub's, weren't measured here. * Snapshot size depends on the page. Long articles and busy pages cost more than short product pages. Data and code#. * Extract Summit 2026: AI writes the scrapers, but the data still needs proof I watched all 11 talks from Extract Summit 2026 in Austin. AI now writes and repairs scrapers in production. Most speakers focused on how to prove the results are right, with verifiers, independent checks and evidence files. 14 October 2026 * What top websites tell AI crawlers: robots.txt rules on 544 sites, bot by bot I checked the robots.txt of 544 top websites for 15 AI and search user agents. None blocks Googlebot, 17% block GPTBot, and many sites treat AI training, AI search and user requests differently. 13 October 2026 * Common Crawl's WET text vs trafilatura vs Resiliparse, retested Hugging Face's FineWeb team found Common Crawl's ready-made text too noisy and extracted their own. I retested that choice on 2,000 pages from the September 2026 crawl, with FineWeb's own quality filter. 8 October 2026

She Code Africa and Apify announce winners of BuildHer Hackathon 2026. Quick Links * Partner She Code Africa and Apify announce winners of BuildHer Hackathon 2026 By Partner 30th Sep, 2026 Get TC updates in your Newsfeed Share She Code Africa, in partnership with global web scraping and data extraction platform Apify, has officially concluded the BuildHer Hackathon 2026, held under the theme "Ship and Earn Africa." The hybrid hackathon brought together women developers and technical builders to create, publish and monetise technology solutions to real-world problems, culminating in an in-person Demo Day on Friday, September 25, at The Zone, Gbagada, Lagos. Aimed at creating more pathways for developers to turn their technical skills into economic opportunities, the hackathon challenged participants to build, deploy and monetise custom Apify Actors using the platform's pay-per-event (PPE) model. Across use cases including AgriData, EduTech, FinTrack, Cybersecurity and SME Tools, teams were evaluated on their technical execution, use of Apify, relevance of their solutions, sustainability and presentation. Speaking about the partnership, Mfonobong Umondia, Community Manager at She Code Africa, said: "The She Code Africa x Apify Hackathon is a hybrid hackathon that brings together women and girls from different engineering tracks to build solutions to real-time problems. So we're basically bringing them together in one room to collaborate and build solutions to problems." Following live evaluations and presentations before a panel of judges, Team Data Divas took first place with WorkDey, Team Byte Queens placed second with the Nigeria Flood Early-Warning Agent, and Team SentinelNG placed third with Sentinel. 1st Place - Team Data Divas: LoanSafe Team Data Divas took the top prize with Loan Safe, an Apify Actor built to help Nigerians verify digital lenders before borrowing. Loan Safe uses Nigeria's official FCCPC lending registry to let users check whether a lender is registered and track changes over time. It stores verified records so users can see when a lender's status changes, a listing disappears, or a new lender appears in the registry. It also gives fintechs, developers and consumer protection researchers a simpler way to access and monitor lender information without building the system from scratch. 2nd Place - Team Byte Queens: Nigeria Flood Early-Warning Agent Team Byte Queens placed second with the Nigeria Flood Early-Warning Agent, a tool designed to turn different sources of flood information into community-level risk alerts. The Actor brings together information from NIHSA flood advisories, GloFAS river-flow forecasts, rainfall forecasts and upstream dam releases. Users can enter a community, Local Government Area or GPS location and receive a flood-risk assessment for the next seven and 30 days, alongside the factors contributing to the assessment and recommended actions. The team also developed a companion WhatsApp bot that delivers alerts in English, Hausa, Yoruba, Igbo and Nigerian Pidgin. The current MVP covers Kogi, Benue, Adamawa, Anambra and Lagos. 3rd Place - Team SentinelNG: Sentinel Taking third place, Team SentinelNG built Sentinel, an external attack-surface monitoring Actor designed for financial institutions and other organisations. Sentinel uses publicly available Certificate Transparency logs to identify internet-facing assets associated with an organisation's domain. It can flag forgotten or unexpected assets, certificate anomalies and potential impersonation domains, and assigns risk scores and severity levels to its findings. In addition to the top three cash prizes, members of the top 10 teams received US$100 in prepaid Apify platform usage credits to support their continued use of the platform. Building beyond technical skills For Apify, the hackathon was also an opportunity to connect with more women developers in Africa and create pathways for developers to monetise their technical skills. Saurav Jain, Senior Developer Community Manager at Apify, said: "With time, we have noticed there is a huge gap in the African market where developers cannot really make money out of their software skills, but they have definitely very good, great developer skills." He said the partnership with She Code Africa was part of Apify's effort to engage more women developer communities, particularly across Africa. "So this year, we made a pledge that we will participate with many women communities all over the world and especially in Africa. So what can be a great start other than collaborating with She Code Africa?" The program will continue beyond Demo Day. The top three winning teams will participate in an exclusive post-hackathon workshop on Wednesday, October 7, 2026, led by Apify's Developer Relations team and She Code Africa. The wider participant community will also continue to receive resources and support after the hackathon. Oluwaseun Oladeji, Community Associate at She Code Africa, said: "After the whole program, of course, we are also going to continuously engage with the participants who participated in this hackathon just to build on their technical skills, as well as improving more on how they can develop their Apify Actors." The post-hackathon engagement is intended to give participants an opportunity to continue developing the tools they built during the programme and strengthen their technical skills beyond the competition. More from this Author * Government Lagos wants its innovation ecosystem to become a global leader * Events TechCabal and Moniepoint partner to discuss how tech is changing the informal economy * Events ARM collaborates with TechCabal on the Future of Commerce partnership pitch session Partner Virtual dollar card vs domiciliary account: which is better for Nigerians in 2026? Partner | 7 minutes ago

She Code Africa and Apify partner to equip African women developers to build and monetise technology for local problems. Published: September 23, 2026 This Brand Press post is for informational purpose only and should not be interpreted as financial or investment guidance. Always ensure to carry out due diligence. Read all... She Code Africa has partnered with Apify to launch the Apify x SCA BuildHer Hackathon 2026, a hybrid programme challenging women developers and technical builders across Africa to create, publish and monetise technology that addresses real-world problems on the continent. Under the theme "Ship and Earn Africa," the hackathon will give participants hands-on experience building their own Apify Actors, web scraping and automation tools hosted on the Apify Store. The programme is designed to take participants beyond building prototypes, requiring teams to develop working tools that can be published and monetised. The partnership brings together the She Code Africa community of women in technology and Apify's web automation and data infrastructure to create a practical pathway into the developer economy. Apify currently hosts more than 50,000 Actors, while its developer community generates more than US$1.5 million monthly, according to the hackathon programme. "At She Code Africa, we want women to have more than access to technology; we want them to have the opportunity to build with it, solve problems that matter to their communities and create economic value from what they build. Our partnership with Apify brings that vision to life by giving women developers a practical opportunity to build, publish and monetise their own tools. Through Ship and Earn Africa, we're creating a space where African women can turn their technical skills into solutions that respond to real challenges across the continent." said Ada Nduka Oyom, Founder & Executive Director of She Code Africa. Participants will build solutions across areas including agricultural intelligence, education, financial data, cybersecurity, jobs, healthcare, legal access, travel, small-business intelligence and data technology. Proposed use cases include monitoring agricultural prices across Nigerian markets, aggregating scholarship opportunities for African students, tracking remote jobs for African developers, and turning fragmented public datasets into useful APIs and data products. The hackathon is targeted at women developers and engineers, AI builders, indie hackers and product builders, as well as students and early-career developers with the coding ability to build and ship a working technical project. Through Apify, participants will have access to free accounts, platform credits, the Apify SDK, learning resources through Apify Academy and technical support from Apify's Developer Relations team throughout the build period. Each participating team will be expected to publish its completed Actor and use Apify's PPE monetisation model. Apify representative quote. "The best thing about the Apify community is that anyone can build a tool, publish it, and start earning from it the same week. African women developers are already solving hard local problems; what's often missing is a path from a working prototype to a product with real users and real income. This hackathon is that path. We're excited to see what the participants ship, and the Apify team will be with them from onboarding through Demo Day and beyond." Saurav Jain, Senior Developer Community Manager, Apify The virtual hackathon onboarding and build session will take place on Thursday, September 24, 2026, followed by an in-person Demo Day and Awards Ceremony on Friday, September 25, 2026, in Lagos. Teams will demonstrate their solutions before a judging panel comprising external judges and Apify representatives, with selected finalists advancing to the final presentation round. The top three teams will receive US$700, US$500 and US$300 respectively, while members of the top 10 teams will receive Apify platform credits. The top three winning teams will also participate in a post-hackathon workshop with Apify's Developer Relations team and She Code Africa to support the continued development of their projects. The Apify x SCA BuildHer Hackathon reflects a shared commitment to creating more pathways for African women to participate in the technology economy as builders and creators, while developing practical digital solutions for challenges within their communities. About She Code Africa. She Code Africa (SCA) is a non-profit organisation creating opportunities for African women to thrive in technology. She Code Africa supports women across different stages of their technology journeys through technology education, mentorship, career acceleration, economic opportunity programmes, community and networks that support their growth within the technology ecosystem. About Apify. Apify is a web automation and data extraction platform that provides developers and businesses with tools and infrastructure to collect web data, automate workflows and build AI-powered applications. Its marketplace, Apify Store, hosts thousands of tools known as Actors, which can be used, customised and integrated into automated workflows. Media contact. She Code Africa Juliana Ajagunna [email protected] Join 30,000 other smart people like you Get our fun 5-minute roundup of happenings in African and global tech, directly in your inbox every weekday, hours before everyone else. More Stories Reach our audience Our publications