--- type: "post" id: "how-we-made-our-docs-test-themselves-analysis-13a388" slug: "how-we-made-our-docs-test-themselves-analysis-13a388" title: "How We Made Our Docs Test Themselves" description: "LangChain’s documentation case shows a division of labor between agent instructions and continuous integration." retrieval_nugget: "LangChain’s documentation case shows a division of labor between agent instructions and continuous integration." published_at: "2026-10-13" updated_at: "2026-10-13" record_date: "2026-10-13" date_kind: "published_at" topics: [] source_urls: ["https://www.langchain.com/blog/our-docs-test-themselves"] schema_version: "newruntime-agent-readable-v0.2" stable_id: "post:how-we-made-our-docs-test-themselves-analysis-13a388" status: "published" visuals: [] editorial_provenance: {"schema_version":"newruntime-editorial-copy-v1","content_status":"source_grounded_final","final_copy_sha256":"sha256:e1dfe973a6834cd13d1330dc4de42149f0d63b1bfe6af727d430256ce6eba6b1","reviewed_at":"2026-10-13T17:07:38.187Z","source_evidence_count":1,"verified_claim_count":2,"site_analysis_schema_version":"newruntime-site-analysis-v1","site_object_kind":"field_note","observed_fact_count":2,"implication_count":2,"watch_condition_count":1,"related_record_count":0} analysis: {"schema_version":"newruntime-site-analysis-v1","object_kind":"field_note","thesis":"Documentation examples become more maintainable when migration into runnable artifacts is paired with recurring execution checks, rather than left as static inline text.","observed_facts":[{"text":"LangChain says its Deep Agents CLI skill moves inline code into standalone files, adds setup and teardown, lints it, marks snippets, runs samples, and regenerates documentation snippets.","source_urls":["https://www.langchain.com/blog/our-docs-test-themselves"]},{"text":"LangChain says a GitHub Action regularly runs the code-sample tests and creates tickets when a test fails.","source_urls":["https://www.langchain.com/blog/our-docs-test-themselves"]}],"mechanism":"A reusable skill encodes the migration procedure, while CI repeatedly tests the resulting artifacts and surfaces failures for follow-up.","why_now":"LangChain says rapid changes to its product, models, SDKs, and practices make stale examples a recurring risk.","implications":["Teams can treat documentation examples as executable assets in the same quality loop as application code.","Agent-assisted migration reduces the manual setup burden, while CI provides recurring detection rather than one-time cleanup."],"evidence_boundary":"This is LangChain’s account of its own implementation, without comparative results or independent validation.","watch_conditions":["Whether LangChain reports that recurring checks are maintained at scale and how often failures lead to corrected documentation rather than merely tickets."],"related_records":[],"new_branch_reason":"No prior New Runtime record is supplied; this article offers a specific implementation to track without assuming continuity with another record."} editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]} routes: {"html":"https://newruntime.com/posts/how-we-made-our-docs-test-themselves-analysis-13a388/","markdown":"https://newruntime.com/posts/how-we-made-our-docs-test-themselves-analysis-13a388.md","json":"https://newruntime.com/posts/how-we-made-our-docs-test-themselves-analysis-13a388.json"} --- # How We Made Our Docs Test Themselves ## Retrieval answer LangChain’s documentation case shows a division of labor between agent instructions and continuous integration. LangChain’s documentation case shows a division of labor between agent instructions and continuous integration. The thesis is: Documentation examples become more maintainable when migration into runnable artifacts is paired with recurring execution checks, rather than left as static inline text. The source reports that Deep Agents CLI used a repository skill to move inline examples into standalone files, add setup and teardown, lint code, mark snippets, run examples, and regenerate documentation snippets. It also says a GitHub Action runs tests regularly and creates tickets on failure. The mechanism is: A reusable skill encodes the migration procedure, while CI repeatedly tests the resulting artifacts and surfaces failures for follow-up. The why_now is: LangChain says rapid changes to its product, models, SDKs, and practices make stale examples a recurring risk. The first implication is: Teams can treat documentation examples as executable assets in the same quality loop as application code. This does not establish that every documentation system can adopt the workflow unchanged, nor does the post quantify reliability gains; the evidence boundary is: This is LangChain’s account of its own implementation, without comparative results or independent validation. The first watch condition is: Whether LangChain reports that recurring checks are maintained at scale and how often failures lead to corrected documentation rather than merely tickets. This starts a distinct analytical record because no prior New Runtime record is supplied; the article offers a specific implementation to track, not a basis for inventing continuity with another record. Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json ## Continue in the graph Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`. - collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis. - collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph. - whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)