{"schema_version":"onlylabs.public_signal.v1","title":"Anthropic Writing: Claude Code Expertise","description":"Anthropic writing signal with public source context, captured evidence pages, related signals, and data-business radar classification.","url":"https://onlylabs.fyi/signals/f8eeaef7-de16-48ed-bf3e-d08eee091d02","json_url":"https://onlylabs.fyi/signals/f8eeaef7-de16-48ed-bf3e-d08eee091d02/signal.json","generated_at":"2026-06-27T00:28:22.601Z","evidence_latest_fetched_at":"2026-06-23T00:04:19.954+00:00","signal_first_seen_at":"2026-06-16T20:01:28.390559+00:00","org":{"slug":"anthropic","name":"Anthropic","category":"frontier-lab","category_label":"Frontier lab","dossier_url":"https://onlylabs.fyi/labs/anthropic","dossier_json_url":"https://onlylabs.fyi/labs/anthropic/dossier.json"},"related_urls":{"signal":"https://onlylabs.fyi/signals/f8eeaef7-de16-48ed-bf3e-d08eee091d02","signal_json":"https://onlylabs.fyi/signals/f8eeaef7-de16-48ed-bf3e-d08eee091d02/signal.json","source":"https://www.anthropic.com/research/claude-code-expertise","lab_dossier":"https://onlylabs.fyi/labs/anthropic","lab_dossier_json":"https://onlylabs.fyi/labs/anthropic/dossier.json","analysis":"https://onlylabs.fyi/analysis/anthropic","analysis_json":"https://onlylabs.fyi/analysis/anthropic/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/anthropic/evidence.json","category":"https://onlylabs.fyi/frontier","category_json":"https://onlylabs.fyi/frontier.json","category_feed":"https://onlylabs.fyi/frontier/feed.xml","category_signals_json":"https://onlylabs.fyi/signals.json","topic":"https://onlylabs.fyi/topics/talking","topic_signals_json":"https://onlylabs.fyi/topics/talking/signals.json","topic_feed":"https://onlylabs.fyi/topics/talking/feed.xml","data_business":null},"answer_pack":{"answer":"Anthropic published Claude Code Expertise. This talking signal gives public context for research themes, product direction, policy, or launch framing. High-signal details: Blog post on Claude's coding capabilities, not a major release. · Agentic coding and persistent returns to expertise Agentic coding and persistent returns to expertise Authors Zoe Hitzig, Maxim Massenkoff, Eva Lyubich, Ryan Heller, and.... onlylabs links this event to 5 captured evidence pages and 6 related writing signals.","signal_desk":"talking","source_context":{"source_url":"https://www.anthropic.com/research/claude-code-expertise","source_host":"anthropic.com","occurred_at":"2026-06-17T00:00:00.000Z","first_seen_at":"2026-06-16T20:01:28.390559+00:00","date_source":"page.visible_date","context":null},"context_markers":[{"label":"Lab","value":"Anthropic","source":"signal"},{"label":"Signal desk","value":"talking","source":"signal"},{"label":"Source host","value":"anthropic.com","source":"source"},{"label":"PDF","value":"linked report","source":"source"},{"label":"Notability","value":"Blog post on Claude's coding capabilities, not a major release.","source":"signal"},{"label":"Watch term","value":"Eval methodology","source":"evidence"},{"label":"Watch term","value":"Data pipeline","source":"evidence"},{"label":"Watch term","value":"Infrastructure","source":"evidence"},{"label":"Watch term","value":"Safety and alignment","source":"evidence"},{"label":"Watch term","value":"Agents and tool use","source":"evidence"}],"evidence_coverage":{"target_pages":5,"captured_pages":5,"readable_pages":5,"capture_methods":["exa","plain"],"missing_page_urls":[],"failed_page_urls":[],"blocked_page_urls":[],"page_urls":["https://www.anthropic.com/research/claude-code-expertise","https://cdn.sanity.io/files/4zrzovbb/website/433472e34b60db1a52ebf0b8c6600f057b6908c5.pdf","https://cdn.sanity.io/files/4zrzovbb/website/a94728142a45694292336165947f8d6e3e1a357e.pdf","https://cdn.sanity.io/files/4zrzovbb/website/7426c33b0e75ab4771c465d30d5bc1019bdd0c9c.pdf","https://suproteem.cc/agents.pdf"],"related_signals":6,"has_source_url":true,"latest_page_fetched_at":"2026-06-23T00:04:19.954+00:00"},"data_business":{"matches":false,"lanes":[],"matched_terms":[],"score":null,"reason":null},"agent_handoff":{"signal_json":"https://onlylabs.fyi/signals/f8eeaef7-de16-48ed-bf3e-d08eee091d02/signal.json","dossier_json":"https://onlylabs.fyi/labs/anthropic/dossier.json","analysis_json":"https://onlylabs.fyi/analysis/anthropic/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/anthropic/evidence.json","topic_signals_json":"https://onlylabs.fyi/topics/talking/signals.json","topic_feed":"https://onlylabs.fyi/topics/talking/feed.xml","category_signals_json":"https://onlylabs.fyi/signals.json","data_radar_json":null,"opportunities_json":null},"analysis_playbook":{"objective":"Turn public writing and discussion into a readable map of research themes, product framing, policy posture, launch narratives, and market attention.","evidence_focus":["post title","source URL","captured page text","HN traction","linked model or paper references","publication date"],"extraction_questions":["Which themes are labs choosing to explain publicly?","Which posts are attracting outside discussion?","Which writing reframes a recent release, model, hiring wave, or policy stance?","Which posts mention data, evals, infrastructure, safety, or deployment workflows?"],"signal_questions":["What public theme, launch framing, or research direction does this writing signal expose?","Which themes are labs choosing to explain publicly?","Which posts are attracting outside discussion?","Do the 6 related writing signals show a repeated pattern?"],"output_fields":["org","theme","public_framing","traction","data_business_lane","evidence_url"],"data_business_relevance":"Public writing supplies the narrative layer over raw signals and helps identify which frontier-lab priorities are becoming externally legible.","required_sources":[{"label":"signal_json","url":"https://onlylabs.fyi/signals/f8eeaef7-de16-48ed-bf3e-d08eee091d02/signal.json","required":true},{"label":"source","url":"https://www.anthropic.com/research/claude-code-expertise","required":true},{"label":"dossier_json","url":"https://onlylabs.fyi/labs/anthropic/dossier.json","required":true},{"label":"analysis_evidence_json","url":"https://onlylabs.fyi/analysis/anthropic/evidence.json","required":true},{"label":"topic_signals_json","url":"https://onlylabs.fyi/topics/talking/signals.json","required":false},{"label":"data_radar_json","url":null,"required":false}],"expected_output":["one-paragraph source-grounded interpretation","category-specific implication","confidence and missing evidence","recommended next source to inspect"],"prompt_seed":"Using only the linked onlylabs JSON, captured source context, and cited evidence, analyze Anthropic's writing signal \"Claude Code Expertise\" for frontier lab strategy."},"semantic_triples":[{"subject":"Anthropic","predicate":"published","object":"Claude Code Expertise","text":"Anthropic published Claude Code Expertise."},{"subject":"Claude Code Expertise","predicate":"is classified as","object":"writing signal","text":"Claude Code Expertise is classified as writing signal."},{"subject":"Claude Code Expertise","predicate":"belongs to","object":"talking desk","text":"Claude Code Expertise belongs to talking desk."},{"subject":"Claude Code Expertise","predicate":"has evidence coverage","object":"5 captured evidence pages","text":"Claude Code Expertise has evidence coverage 5 captured evidence pages."},{"subject":"Claude Code Expertise","predicate":"has captured page count","object":"5","text":"Claude Code Expertise has captured page count 5."},{"subject":"Claude Code Expertise","predicate":"has readable page count","object":"5","text":"Claude Code Expertise has readable page count 5."},{"subject":"Claude Code Expertise","predicate":"has related signal count","object":"6","text":"Claude Code Expertise has related signal count 6."},{"subject":"Claude Code Expertise","predicate":"has analysis playbook objective","object":"Turn public writing and discussion into a readable map of research themes, product framing, policy posture, launch narratives, and market attention.","text":"Claude Code Expertise has analysis playbook objective Turn public writing and discussion into a readable map of research themes, product framing, policy posture, launch narratives, and market attention.."},{"subject":"Claude Code Expertise","predicate":"has source host","object":"anthropic.com","text":"Claude Code Expertise has source host anthropic.com."},{"subject":"Claude Code Expertise","predicate":"has lab","object":"Anthropic","text":"Claude Code Expertise has lab Anthropic."},{"subject":"Claude Code Expertise","predicate":"has signal desk","object":"talking","text":"Claude Code Expertise has signal desk talking."},{"subject":"Claude Code Expertise","predicate":"has source host","object":"anthropic.com","text":"Claude Code Expertise has source host anthropic.com."},{"subject":"Claude Code Expertise","predicate":"has pdf","object":"linked report","text":"Claude Code Expertise has pdf linked report."},{"subject":"Claude Code Expertise","predicate":"has notability","object":"Blog post on Claude's coding capabilities, not a major release.","text":"Claude Code Expertise has notability Blog post on Claude's coding capabilities, not a major release.."},{"subject":"Claude Code Expertise","predicate":"has watch term","object":"Eval methodology","text":"Claude Code Expertise has watch term Eval methodology."},{"subject":"Claude Code Expertise","predicate":"has watch term","object":"Data pipeline","text":"Claude Code Expertise has watch term Data pipeline."},{"subject":"Claude Code Expertise","predicate":"has watch term","object":"Infrastructure","text":"Claude Code Expertise has watch term Infrastructure."}]},"intelligence":{"signal_desk":"talking","answer":"Anthropic published Claude Code Expertise. This talking signal gives public context for research themes, product direction, policy, or launch framing. High-signal details: Blog post on Claude's coding capabilities, not a major release. · Agentic coding and persistent returns to expertise Agentic coding and persistent returns to expertise Authors Zoe Hitzig, Maxim Massenkoff, Eva Lyubich, Ryan Heller, and.... onlylabs links this event to 5 captured evidence pages and 6 related writing signals.","semantic_triples":[{"subject":"Anthropic","predicate":"published","object":"Claude Code Expertise","text":"Anthropic published Claude Code Expertise."},{"subject":"Claude Code Expertise","predicate":"is classified as","object":"writing signal","text":"Claude Code Expertise is classified as writing signal."},{"subject":"Claude Code Expertise","predicate":"belongs to","object":"talking desk","text":"Claude Code Expertise belongs to talking desk."},{"subject":"Claude Code Expertise","predicate":"has evidence coverage","object":"5 captured evidence pages","text":"Claude Code Expertise has evidence coverage 5 captured evidence pages."}]},"signal":{"id":"f8eeaef7-de16-48ed-bf3e-d08eee091d02","url":"https://onlylabs.fyi/signals/f8eeaef7-de16-48ed-bf3e-d08eee091d02","json_url":"https://onlylabs.fyi/signals/f8eeaef7-de16-48ed-bf3e-d08eee091d02/signal.json","source_url":"https://www.anthropic.com/research/claude-code-expertise","title":"Claude Code Expertise","summary":"Anthropic published a writing signal. onlylabs watches public writing for research themes, product direction, and model-launch context.","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"anthropic","name":"Anthropic","category":"frontier-lab"},"occurred_at":"2026-06-17T00:00:00.000Z","first_seen_at":"2026-06-16T20:01:28.390559+00:00","date_source":"page.visible_date","evidence_coverage":{"target_pages":5,"captured_pages":5,"readable_pages":5,"capture_methods":["exa","plain"],"missing_page_urls":[],"failed_page_urls":[],"blocked_page_urls":[],"page_urls":["https://www.anthropic.com/research/claude-code-expertise","https://cdn.sanity.io/files/4zrzovbb/website/433472e34b60db1a52ebf0b8c6600f057b6908c5.pdf","https://cdn.sanity.io/files/4zrzovbb/website/a94728142a45694292336165947f8d6e3e1a357e.pdf","https://cdn.sanity.io/files/4zrzovbb/website/7426c33b0e75ab4771c465d30d5bc1019bdd0c9c.pdf","https://suproteem.cc/agents.pdf"]},"facets":{},"traction":{"github_stars":null,"hn_points":null,"hn_comments":null,"hn_story_id":null,"hf_downloads":null,"hf_likes":null},"data_radar":null},"primary_evidence_page":{"is_primary":true,"source_match":true,"url":"https://www.anthropic.com/research/claude-code-expertise","final_url":"https://www.anthropic.com/research/claude-code-expertise","title":"Claude Code Expertise","http_status":200,"content_type":"text/html; charset=utf-8","capture_method":"plain","fetched_at":"2026-06-16T20:04:06.419043+00:00","bytes":219178,"raw_path":"a6ebb163a6c3bf175c39e54eeac5010fd84b9e52c0a6e23dcf5da992c57fed9e.html","content_hash":"f3c0c818132e7a3efad64093545a96a8fe9838d2a6e3b18036c19497ba0d91ef","excerpt_chars":1200,"truncated":true,"excerpt":"Agentic coding and persistent returns to expertise \\ Anthropic Economic Research Agentic coding and persistent returns to expertise Jun 17, 2026 Read in PDF Key findings Building on prior work , we introduce a framework for studying interactive agentic coding based on a privacy-preserving analysis of ~400,000 Claude Code sessions from between October 2025 and April 2026. We evaluate the composition of tasks, human-AI collaboration, and success rates. In a typical session, people make most of the planning decisions (what to do) and Claude makes most of the execution decisions (how to do it). The greater domain expertise a person brings to a session, the more work Claude does per instruction. On coding tasks, every major occupation succeeds––accomplishes what the person set out to do, with verifiable evidence like passing tests or committed work––at nearly the same rate as software engineers, on average. The more domain expertise a person has, the more often the session ends in success—though the gap between intermediate and expert users is modest. Over the seven months we observe, the share of sessions spent debugging fell by nearly half, and usage shifted toward more end-to-end..."},"evidence_pages":[{"is_primary":false,"source_match":false,"url":"https://cdn.sanity.io/files/4zrzovbb/website/433472e34b60db1a52ebf0b8c6600f057b6908c5.pdf","final_url":"https://cdn.sanity.io/files/4zrzovbb/website/433472e34b60db1a52ebf0b8c6600f057b6908c5.pdf","title":"Claude Code Expertise","http_status":200,"content_type":"application/pdf","capture_method":"exa","fetched_at":"2026-06-23T00:04:19.954+00:00","bytes":1911775,"raw_path":"35f28ad8c49a24e3abf93f0b89211b9d4e36ac9591387bd12bf99891643aa85f.pdf","content_hash":"644d0e3fe30c350d10217f4db0051f4e36bc4cceaa3e5d27cd3efdcee2e444f0","excerpt_chars":1200,"truncated":true,"excerpt":"Agentic coding and persistent returns to expertise Agentic coding and persistent returns to expertise Authors Zoe Hitzig, Maxim Massenkoff, Eva Lyubich, Ryan Heller, and Peter McCrory Published June 16, 2026 Acknowledgements With acknowledgements to: Jake Eaton, Sarah Pollack, Hanah Ho, Szymon Sacher, Anton Korinek, Santi Ruiz, Kerry Persen, Ankur Rathi, Alex Tamkin, Heather Whitney, Cat Wu, Kacie Jenkins, Jennifer Martinez, Amie Rotherham, Boris Cherny, Eleanor Dorfman, Miles McCain, and Jack Clark. Key findings • Building on prior work, we introduce a framework for studying interactive agentic coding based on a privacy-preserving analysis of ~400,000 Claude Code sessions from between October 2025 and April 2026. We evaluate the composition of tasks, human-AI collaboration, and success rates. • In a typical session, people make most of the planning decisions (what to do) and Claude makes most of the execution decisions (how to do it). The greater domain expertise a person brings to a session, the more work Claude does per instruction. On coding tasks, every major occupation succeeds—accomplishes what the person set out to do, with verifiable evidence like passing tests or..."},{"is_primary":false,"source_match":false,"url":"https://cdn.sanity.io/files/4zrzovbb/website/a94728142a45694292336165947f8d6e3e1a357e.pdf","final_url":"https://cdn.sanity.io/files/4zrzovbb/website/a94728142a45694292336165947f8d6e3e1a357e.pdf","title":"Claude Code Expertise","http_status":200,"content_type":"application/pdf","capture_method":"exa","fetched_at":"2026-06-23T00:04:10.173+00:00","bytes":560586,"raw_path":"731276d57c26583f9ce899ef66779e810ec2b4fb581386a9c37b5cbd1c46ddeb.pdf","content_hash":"43139a8128eddfadc8d739a0d1faa3550a65057a730fa3cdc5e536ce8f7b2e8a","excerpt_chars":1200,"truncated":true,"excerpt":"anthropic.com Appendix to “Agentic Coding and Persistent Returns to Expertise” June 2026 Sample Our sample covers interactive Claude Code sessions between October 2025 and April 2026 (inclusive). This includes sessions conducted through the Claude Code command-line interface, the Claude Code desktop app, and claude.ai. This excludes Claude Code usage via third party integrated developer environments as well as via software development kits. Sessions with zero human turns are excluded—these excluded sessions are non interactive calls in “headless” mode i.e., via claude -p “ ”. The analysis covers 398,198 sessions uniformly randomly sampled from Claude Code traffic. These sessions come from 234,751 users. 73,066 users appear more than once in the sample, 6,382 users appear more than 5 times and 1,413 users appear more than 10 times. We exclude internal Anthropic usage. All classifications ran through our privacy-preserving analysis infrastructure. Researchers never read individual sessions or transcripts, and every number in this report is an aggregate over a cell of sessions that is suppressed unless it contains a minimum number of distinct users. For each classifier call, the..."},{"is_primary":false,"source_match":false,"url":"https://cdn.sanity.io/files/4zrzovbb/website/7426c33b0e75ab4771c465d30d5bc1019bdd0c9c.pdf","final_url":"https://cdn.sanity.io/files/4zrzovbb/website/7426c33b0e75ab4771c465d30d5bc1019bdd0c9c.pdf","title":"Claude Code Expertise","http_status":200,"content_type":"application/pdf","capture_method":"exa","fetched_at":"2026-06-23T00:04:09.572+00:00","bytes":560584,"raw_path":"2494c5f4c00dc3b01a8def85af7671a5642e2977c522adb7555aaa392c8ed3e3.pdf","content_hash":"8b34631d4f5e21526d42e986cca168656c378a4711277cc67f6131384725b934","excerpt_chars":1200,"truncated":true,"excerpt":"anthropic.com Appendix to “Agentic Coding and Persistent Returns to Expertise” June 2026 Sample Our sample covers interactive Claude Code sessions between October 2025 and April 2026 (inclusive). This includes sessions conducted through the Claude Code command-line interface, the Claude Code desktop app, and claude.ai. This excludes Claude Code usage via third party integrated developer environments as well as via software development kits. Sessions with zero human turns are excluded—these excluded sessions are non interactive calls in “headless” mode i.e., via claude -p “ ”. The analysis covers 398,198 sessions uniformly randomly sampled from Claude Code traffic. These sessions come from 234,751 users. 73,066 users appear more than once in the sample, 6,382 users appear more than 5 times and 1,413 users appear more than 10 times. We exclude internal Anthropic usage. All classifications ran through our privacy-preserving analysis infrastructure.. Researchers never read individual sessions or transcripts, and every number in this report is an aggregate over a cell of sessions that is suppressed unless it contains a minimum number of distinct users. For each classifier call, the..."},{"is_primary":false,"source_match":false,"url":"https://suproteem.cc/agents.pdf","final_url":"https://suproteem.is/assets/files/agents.pdf","title":"Claude Code Expertise","http_status":200,"content_type":"application/pdf","capture_method":"exa","fetched_at":"2026-06-16T20:04:13.814+00:00","bytes":5029006,"raw_path":"ab9084ede7fbc332fe837190412d49f7cc2705f91cea6d6c23e29f904ea7e1a1.pdf","content_hash":"0c2825e06a26e74a2171ab26add3487ee9319f94351a0849c6e0b928be74df55","excerpt_chars":1200,"truncated":true,"excerpt":"AI Agents and Higher-Order Work Suproteem K. Sarkar University of Chicago May 8, 2026 Please click here for the latest version How do AI agents influence knowledge work? This paper finds that agents shift worker effort from implementation to supervision, which especially benefits verifiable work and expert workers. I analyze data from the coding platform Cursor to study agents in software production. First, I find that workers have become less likely to produce output manually after AI agents were introduced, and are more likely to delegate work to agents. Second, workers use agents for abstract, higher-order tasks like delegation, context gathering, and planning. Third, agents are used more frequently in settings with easier-to-verify work outputs. Fourth, experienced workers appear more skilled at delegating work to agents—they ask fewer questions, seek more alignment through planning, and accept agent outputs at higher rates. Fifth, using variation from the platform’s feature release timeline, I find that agents increase software output, especially for firms with more verifiable work and more experienced workers. The complementarity between AI agents and expertise contrasts..."}],"related_signals":[{"id":"9caf7d39-0d96-42d0-8b59-e72f7e502ec4","url":"https://onlylabs.fyi/signals/9caf7d39-0d96-42d0-8b59-e72f7e502ec4","source_url":"https://www.anthropic.com/research/economic-index-june-2026-report","title":"Economic Index June 2026 Report","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"anthropic","name":"Anthropic","category":"frontier-lab"},"occurred_at":"2026-06-26T23:50:17+00:00","first_seen_at":"2026-06-26T16:00:32.297045+00:00","date_source":"sitemap.lastmod"},{"id":"6c78c028-3ab4-4b33-86f7-d86c8ba9e3ba","url":"https://onlylabs.fyi/signals/6c78c028-3ab4-4b33-86f7-d86c8ba9e3ba","source_url":"https://www.anthropic.com/research/agents-in-biology","title":"Agents In Biology","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"anthropic","name":"Anthropic","category":"frontier-lab"},"occurred_at":"2026-06-08T00:00:00.000Z","first_seen_at":"2026-06-09T02:17:26.339488+00:00","date_source":"page.visible_date"},{"id":"e4fbfcdd-15b4-41b9-b011-fd83e7068ae9","url":"https://onlylabs.fyi/signals/e4fbfcdd-15b4-41b9-b011-fd83e7068ae9","source_url":"https://www.anthropic.com/research/making-claude-a-chemist","title":"Making Claude A Chemist","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"anthropic","name":"Anthropic","category":"frontier-lab"},"occurred_at":"2026-06-05T00:00:00.000Z","first_seen_at":"2026-06-09T02:17:26.339488+00:00","date_source":"page.visible_date"},{"id":"b8472bfd-c32b-4d72-ae98-d8a4f1a1278f","url":"https://onlylabs.fyi/signals/b8472bfd-c32b-4d72-ae98-d8a4f1a1278f","source_url":"https://www.anthropic.com/research/exploit","title":"Exploit","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"anthropic","name":"Anthropic","category":"frontier-lab"},"occurred_at":"2026-03-06T00:00:00.000Z","first_seen_at":"2026-06-17T20:00:31.646229+00:00","date_source":"page.visible_date"},{"id":"5f0f346c-0718-4d81-a174-d18cae51efa3","url":"https://onlylabs.fyi/signals/5f0f346c-0718-4d81-a174-d18cae51efa3","source_url":"https://www.anthropic.com/research/critical-infrastructure-defense","title":"Critical Infrastructure Defense","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"anthropic","name":"Anthropic","category":"frontier-lab"},"occurred_at":"2026-01-08T00:00:00.000Z","first_seen_at":"2026-06-17T20:00:31.646229+00:00","date_source":"page.visible_date"},{"id":"b8ad5d26-1905-4d23-a2e0-c2c477cbb462","url":"https://onlylabs.fyi/signals/b8ad5d26-1905-4d23-a2e0-c2c477cbb462","source_url":"https://www.anthropic.com/news/core-views-on-ai-safety","title":"Core Views On Ai Safety","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"anthropic","name":"Anthropic","category":"frontier-lab"},"occurred_at":"2023-03-08T00:00:00.000Z","first_seen_at":"2026-06-09T02:17:26.339488+00:00","date_source":"page.visible_date"}]}