{"schemaVersion":1,"id":"https://lucioduran.com/","url":"https://lucioduran.com/","title":"Lucio Durán","description":"22+ years architecting and shipping software products across healthcare, e-commerce, and marketplace verticals, from system design to production deployment.\nI lead engineering organizations where AI is not a feature but the core infrastructure layer. My work spans agentic AI systems, LLM integration patterns, full-stack development (React, React Native, Node, Python, iOS), and the strategic decisions that turn technical capabilities into revenue.\nI've built and scaled cross-functional teams across LATAM, the US, and Europe, owning the full product lifecycle: architecture decisions, CI/CD pipelines, cloud infrastructure, API design, and frontend performance. I operate at the intersection of hands-on engineering and executive technology strategy, I still read the PRs, and I still make the technical calls.\nCurrent focus: AI-augmented product development, agentic workflows, and building technology that compounds.","language":"en","datePublished":null,"dateModified":null,"contentFormat":"text/markdown","content":"# Lucio Durán\n\nTechnology director and product architect\n\nChief Technology Officer · LOi · San José de Mayo · Uruguay\n\n22+ years architecting and shipping software products across healthcare, e-commerce, and marketplace verticals, from system design to production deployment. I lead engineering organizations where AI is not a feature but the core infrastructure layer. My work spans agentic AI systems, LLM integration patterns, full-stack development (React, React Native, Node, Python, iOS), and the strategic decisions that turn technical capabilities into revenue. I've built and scaled cross-functional teams across LATAM, the US, and Europe, owning the full product lifecycle: architecture decisions, CI/CD pipelines, cloud infrastructure, API design, and frontend performance. I operate at the intersection of hands-on engineering and executive technology strategy, I still read the PRs, and I still make the technical calls. Current focus: AI-augmented product development, agentic workflows, and building technology that compounds.\n\n## Featured projects\n\n- ### [Citable Publication]()\n\n Deterministic HTML, Markdown and JSON publishing from one corpus, with integrity verification before release and reproducible faults.\n\n Python · License: MIT\n\n [Read the project]() · [Repository]()\n- ### [RAG Evaluation Workbench]()\n\n Reproducible RAG evaluation with BM25, dense and hybrid retrieval, local CPU inference, and separate retrieval and answer metrics.\n\n Python · License: MIT\n\n [Read the project]() · [Repository]()\n- ### [Agent Replay Runtime]()\n\n Durable tool execution with event replay, idempotency keys and recovery after real process crashes. Includes effect traces and cases that stop instead of retrying automatically.\n\n Python · License: MIT\n\n [Read the project]() · [Repository]()\n- ### [Agent Injection Evals]()\n\n Reproducible prompt-injection evaluation with a pinned local model, simulated tools and separate metrics for task capability, attack attempts and realized effects.\n\n Python · License: MIT\n\n [Read the project]() · [Repository]()\n- ### [MCP Contract Harness]()\n\n Read-only MCP server and independent clients for checking stdio negotiation, pagination, errors and cancellation, with actual wire messages and reproducible reports.\n\n Python / JavaScript · License: MIT\n\n [Read the project]() · [Repository]()\n\n[All projects]()\n\n## Contents\n\n- ### [Experience]()\n\n Roles, companies, responsibilities and contributions.\n- ### [Expertise]()\n\n Areas of work, experience and related technologies, with individual records and references.\n- ### [Education]()\n\n Courses, institutions and credential links.\n- ### [Biography]()\n\n Career and personal history.\n- ### [Blog]()\n\n Technical publications, ordered by date.\n- ### [Projects]()\n\n Public software, source code and reproducible results.\n- ### [Architecture decisions]()\n\n Architectural context, decisions and consequences.\n- ### [CV]()\n\n Complete curriculum vitae, with print support.\n- ### [Contact]()\n\n Contact information and public profiles.\n- ### [Complete index]()\n\n Every page, including individual records.\n- ### [Content access]()\n\n Formats, sources and how this site works.\n- ### [Privacy]()\n\n Privacy policy.\n\n## Links\n\n- [email@lucioduran.com]()\n- [Github]()\n- [Facebook]()\n- [LinkedIn]()\n- [Twitter]()\n- [Instagram]()\n\n## Built for machine reading\n\nDiscover. Retrieve. Understand. Cite. This infrastructure makes my work accessible as documents and data as well as web pages.\n\n44 implemented controls / 8 areas / EN + ES\n\nEach mark describes an implementation and links to an interface or its documentation. This is a technical site inventory, not external certification or a promise of inclusion in AI answers.\n\n[JSON checklist]() · [Build inventory and hashes]() · [Contracts and examples]()\n\n### 01 / Discovery\n\n- Implemented: D01\n\n **Public crawling**\n\n robots.txt allows public content and advertises the sitemap. [robots.txt]()\n- Implemented: D02\n\n **XML publication map**\n\n The sitemap is built from the published catalog, including individual records. [Sitemap]()\n- Implemented: D03\n\n **Index without JavaScript**\n\n Every page has an HTML link in an index organized by content kind. [HTML]()\n- Implemented: D04\n\n **Paginated JSON catalog**\n\n Canonical URLs, kinds, languages, available dates and representation links. [HTTP / JSON]()\n- Implemented: D05\n\n **llms.txt guide**\n\n Index for clients using this proposal. Generated from the same catalog. [llms.txt]()\n- Implemented: D06\n\n **Discovery on every page**\n\n HTML and Link headers advertise alternate formats and the API catalog. [HTML / Link]()\n\n### 02 / Content and representations\n\n- Implemented: F01\n\n **Server-delivered content**\n\n Main text is in the initial HTML response, before JavaScript runs. [HTML]()\n- Implemented: F02\n\n **Markdown by URL**\n\n Every page supports .md. The English root uses /index.md. [Markdown]()\n- Implemented: F03\n\n **JSON by URL**\n\n Content, headings, sections, links, language and structured data in one document. [HTTP / JSON]()\n- Implemented: F04\n\n **HTTP content negotiation**\n\n Accept selects explicit Markdown or JSON; Vary: Accept separates representations. [Accept / Vary]()\n- Implemented: F05\n\n **One source for three formats**\n\n Markdown and JSON derive from HTML main, without a separate biography for bots. [HTTP / JSON]()\n- Implemented: F06\n\n **Structure preserved**\n\n Lists, tables, code, links and image references survive serialization. [Markdown]()\n- Implemented: F07\n\n **Decoration excluded from content**\n\n Navigation, controls and decorative seals are excluded from Markdown and JSON. [HTTP / JSON]()\n\n### 03 / Identity, attribution and citations\n\n- Implemented: C01\n\n **Canonical document URLs**\n\n HTML, Markdown, JSON and search results refer to the same publication. [HTTP / JSON]()\n- Implemented: C02\n\n **Anchored sections**\n\n Headings have identifiers for linking a specific passage. [HTTP / JSON]()\n- Implemented: C03\n\n **Consistent identity**\n\n Person uses a stable identifier and links to the author’s public profiles. [JSON-LD]()\n- Implemented: C04\n\n **Content-specific structured data**\n\n Profile, articles, projects and credentials publish structured data alongside visible text. [JSON-LD]()\n- Implemented: C05\n\n **Declared author and dates**\n\n Articles identify the author and preserve available publication and modification dates. [JSON Feed]()\n- Implemented: C06\n\n **Project evidence and limitations**\n\n Projects link repositories, revisions and published results with their scope. [Fuentes / Sources]()\n\n### 04 / Retrieval and context\n\n- Implemented: R01\n\n **Search over published text**\n\n BM25 searches passages extracted from build HTML. Scores measure lexical relevance. [HTTP / JSON]()\n- Implemented: R02\n\n **Explicit filters and pagination**\n\n Language, kind, offset, limit and next support bounded, repeatable traversal. [HTTP / JSON]()\n- Implemented: R03\n\n **Results with citations and versions**\n\n Every result includes an excerpt, citationUrl and the indexed document’s contentHash. [HTTP / JSON]()\n- Implemented: R04\n\n **Read individual sections**\n\n section selects a heading and its subsections, with its own hash and source hash. [HTTP / JSON]()\n- Implemented: R05\n\n **MCP context windows**\n\n Reading supports limits and continuation by Unicode characters. The hash identifies the complete source. [MCP]()\n\n### 05 / Contracts and interfaces\n\n- Implemented: I01\n\n **RFC 9727 API catalog**\n\n A Linkset advertises endpoints and links their contracts and documentation. [Linkset]()\n- Implemented: I02\n\n **OpenAPI 3.1 contract**\n\n Operations, parameters, limits, responses and errors are documented for HTTP clients. [OpenAPI]()\n- Implemented: I03\n\n **Document and passage JSON Schema**\n\n The two response types have distinct schemas and a describedby header. [JSON Schema]()\n- Implemented: I04\n\n **Read-only MCP**\n\n Tools to list, search and read using the official SDK and HTTP transport without persistent sessions. [MCP / HTTP]()\n- Implemented: I05\n\n **RFC 9457 structured errors**\n\n Document, search, catalog, inventory and feed APIs return problem+json for handled errors. [Problem Details]()\n\n### 06 / Freshness and integrity\n\n- Implemented: V01\n\n **SHA-256 content identity**\n\n Documents explain which fields are hashed. It identifies a version, not the truth of its claims. [HTTP / JSON]()\n- Implemented: V02\n\n **Build inventory**\n\n Paginated URLs and hashes, collection version and generation time for comparing snapshots. [HTTP / JSON]()\n- Implemented: V03\n\n **Conditional reads**\n\n ETag and If-None-Match allow 304 responses without retransmitting unchanged documents. [ETag / 304]()\n- Implemented: V04\n\n **Metadata without a response body**\n\n HEAD is available on documents, search, catalog, inventory and feeds. [HEAD]()\n- Implemented: V05\n\n **Publication feeds**\n\n JSON Feed and Atom include article text; RSS publishes summaries. Dates and links come from source files. [JSON Feed / Atom / RSS]()\n\n### 07 / Reading and access\n\n- Implemented: A01\n\n **Explicit languages**\n\n HTML declares the content language; JSON and document responses preserve it. [HTTP / JSON]()\n- Implemented: A02\n\n **Linked EN and ES versions**\n\n hreflang links available versions and declares x-default. [HTML head]()\n- Implemented: A03\n\n **Native HTML semantics**\n\n main, headings, lists, tables and links express the content structure. [HTML]()\n- Implemented: A04\n\n **Photo references, decorative seals**\n\n Photos are retained. Project seals are empty elements with aria-hidden. [HTML / aria-hidden]()\n- Implemented: A05\n\n **Public APIs without sessions**\n\n HTTP read APIs support CORS and OPTIONS. No account, cookies or key required. [CORS / OPTIONS]()\n\n### 08 / Consistency and boundaries\n\n- Implemented: B01\n\n **Versioned profile**\n\n Pages and APIs use the same versioned professional profile snapshot. [HTTP / JSON]()\n- Implemented: B02\n\n **Index generated from HTML**\n\n The build rejects missing or duplicate heading anchors and expected pages absent from the index. [Build]()\n- Implemented: B03\n\n **Site-scoped reading**\n\n The reader accepts known public paths and bounds source size, time and redirects. [Contrato / Contract]()\n- Implemented: B04\n\n **Content treated as source material**\n\n MCP tools identify document text as untrusted material for agent instructions. [MCP instructions]()\n- Implemented: B05\n\n **Machine-readable checklist**\n\n This list and its JSON use the same registry. Each control has an identifier and evidence link. [HTTP / JSON]()\n\n### Standards, proposals and scope\n\nHTTP, HTML, JSON Schema, OpenAPI, feeds and MCP solve specific tasks. llms.txt is a discovery proposal. Custom manifests are identified as such. WebMCP remains a developing specification and this site does not claim compatibility with it. Future capabilities will be added with a verifiable contract once they exist.\n\n[Google AI features]() · [OpenAI crawlers]() · [MCP]() · [llms.txt]() · [WebMCP]()","headings":[{"level":1,"text":"Lucio Durán","url":"https://lucioduran.com/#title"},{"level":2,"text":"Featured projects","url":"https://lucioduran.com/#featured-projects"},{"level":3,"text":"Citable Publication","url":"https://lucioduran.com/#project-citable-publication"},{"level":3,"text":"RAG Evaluation Workbench","url":"https://lucioduran.com/#project-rag-evaluation-workbench"},{"level":3,"text":"Agent Replay Runtime","url":"https://lucioduran.com/#project-agent-replay-runtime"},{"level":3,"text":"Agent Injection Evals","url":"https://lucioduran.com/#project-agent-injection-evals"},{"level":3,"text":"MCP Contract Harness","url":"https://lucioduran.com/#project-mcp-contract-harness"},{"level":2,"text":"Contents","url":"https://lucioduran.com/#contents"},{"level":3,"text":"Experience","url":"https://lucioduran.com/#section-experience"},{"level":3,"text":"Expertise","url":"https://lucioduran.com/#section-expertise"},{"level":3,"text":"Education","url":"https://lucioduran.com/#section-education"},{"level":3,"text":"Biography","url":"https://lucioduran.com/#section-bio"},{"level":3,"text":"Blog","url":"https://lucioduran.com/#section-blog"},{"level":3,"text":"Projects","url":"https://lucioduran.com/#section-projects"},{"level":3,"text":"Architecture decisions","url":"https://lucioduran.com/#section-decisions"},{"level":3,"text":"CV","url":"https://lucioduran.com/#section-cv"},{"level":3,"text":"Contact","url":"https://lucioduran.com/#section-contact"},{"level":3,"text":"Complete index","url":"https://lucioduran.com/#section-index"},{"level":3,"text":"Content access","url":"https://lucioduran.com/#section-access"},{"level":3,"text":"Privacy","url":"https://lucioduran.com/#section-privacy"},{"level":2,"text":"Links","url":"https://lucioduran.com/#links"},{"level":2,"text":"Built for machine reading","url":"https://lucioduran.com/#agent-readiness"},{"level":3,"text":"01 / Discovery","url":"https://lucioduran.com/#cap-discovery"},{"level":3,"text":"02 / Content and representations","url":"https://lucioduran.com/#cap-representations"},{"level":3,"text":"03 / Identity, attribution and citations","url":"https://lucioduran.com/#cap-citations"},{"level":3,"text":"04 / Retrieval and context","url":"https://lucioduran.com/#cap-retrieval"},{"level":3,"text":"05 / Contracts and interfaces","url":"https://lucioduran.com/#cap-interfaces"},{"level":3,"text":"06 / Freshness and integrity","url":"https://lucioduran.com/#cap-freshness"},{"level":3,"text":"07 / Reading and access","url":"https://lucioduran.com/#cap-accessibility"},{"level":3,"text":"08 / Consistency and boundaries","url":"https://lucioduran.com/#cap-boundaries"},{"level":3,"text":"Standards, proposals and scope","url":"https://lucioduran.com/#agent-readiness-scope"}],"sections":[{"heading":"Lucio Durán","level":1,"url":"https://lucioduran.com/","content":""},{"heading":"Lucio Durán","level":1,"url":"https://lucioduran.com/#title","content":"Technology director and product architect\n\nChief Technology Officer · LOi · San José de Mayo · Uruguay\n\n22+ years architecting and shipping software products across healthcare, e-commerce, and marketplace verticals, from system design to production deployment. I lead engineering organizations where AI is not a feature but the core infrastructure layer. My work spans agentic AI systems, LLM integration patterns, full-stack development (React, React Native, Node, Python, iOS), and the strategic decisions that turn technical capabilities into revenue. I've built and scaled cross-functional teams across LATAM, the US, and Europe, owning the full product lifecycle: architecture decisions, CI/CD pipelines, cloud infrastructure, API design, and frontend performance. I operate at the intersection of hands-on engineering and executive technology strategy, I still read the PRs, and I still make the technical calls. Current focus: AI-augmented product development, agentic workflows, and building technology that compounds."},{"heading":"Featured projects","level":2,"url":"https://lucioduran.com/#featured-projects","content":""},{"heading":"Citable Publication","level":3,"url":"https://lucioduran.com/#project-citable-publication","content":"Deterministic HTML, Markdown and JSON publishing from one corpus, with integrity verification before release and reproducible faults.\n\nPython · License: MIT\n\n[Read the project]() · [Repository]()"},{"heading":"RAG Evaluation Workbench","level":3,"url":"https://lucioduran.com/#project-rag-evaluation-workbench","content":"Reproducible RAG evaluation with BM25, dense and hybrid retrieval, local CPU inference, and separate retrieval and answer metrics.\n\nPython · License: MIT\n\n[Read the project]() · [Repository]()"},{"heading":"Agent Replay Runtime","level":3,"url":"https://lucioduran.com/#project-agent-replay-runtime","content":"Durable tool execution with event replay, idempotency keys and recovery after real process crashes. Includes effect traces and cases that stop instead of retrying automatically.\n\nPython · License: MIT\n\n[Read the project]() · [Repository]()"},{"heading":"Agent Injection Evals","level":3,"url":"https://lucioduran.com/#project-agent-injection-evals","content":"Reproducible prompt-injection evaluation with a pinned local model, simulated tools and separate metrics for task capability, attack attempts and realized effects.\n\nPython · License: MIT\n\n[Read the project]() · [Repository]()"},{"heading":"MCP Contract Harness","level":3,"url":"https://lucioduran.com/#project-mcp-contract-harness","content":"Read-only MCP server and independent clients for checking stdio negotiation, pagination, errors and cancellation, with actual wire messages and reproducible reports.\n\nPython / JavaScript · License: MIT\n\n[Read the project]() · [Repository]()\n\n[All projects]()"},{"heading":"Contents","level":2,"url":"https://lucioduran.com/#contents","content":""},{"heading":"Experience","level":3,"url":"https://lucioduran.com/#section-experience","content":"Roles, companies, responsibilities and contributions."},{"heading":"Expertise","level":3,"url":"https://lucioduran.com/#section-expertise","content":"Areas of work, experience and related technologies, with individual records and references."},{"heading":"Education","level":3,"url":"https://lucioduran.com/#section-education","content":"Courses, institutions and credential links."},{"heading":"Biography","level":3,"url":"https://lucioduran.com/#section-bio","content":"Career and personal history."},{"heading":"Blog","level":3,"url":"https://lucioduran.com/#section-blog","content":"Technical publications, ordered by date."},{"heading":"Projects","level":3,"url":"https://lucioduran.com/#section-projects","content":"Public software, source code and reproducible results."},{"heading":"Architecture decisions","level":3,"url":"https://lucioduran.com/#section-decisions","content":"Architectural context, decisions and consequences."},{"heading":"CV","level":3,"url":"https://lucioduran.com/#section-cv","content":"Complete curriculum vitae, with print support."},{"heading":"Contact","level":3,"url":"https://lucioduran.com/#section-contact","content":"Contact information and public profiles."},{"heading":"Complete index","level":3,"url":"https://lucioduran.com/#section-index","content":"Every page, including individual records."},{"heading":"Content access","level":3,"url":"https://lucioduran.com/#section-access","content":"Formats, sources and how this site works."},{"heading":"Privacy","level":3,"url":"https://lucioduran.com/#section-privacy","content":"Privacy policy."},{"heading":"Links","level":2,"url":"https://lucioduran.com/#links","content":"- [email@lucioduran.com]()\n- [Github]()\n- [Facebook]()\n- [LinkedIn]()\n- [Twitter]()\n- [Instagram]()"},{"heading":"Built for machine reading","level":2,"url":"https://lucioduran.com/#agent-readiness","content":"Discover. Retrieve. Understand. Cite. This infrastructure makes my work accessible as documents and data as well as web pages.\n\n44 implemented controls / 8 areas / EN + ES\n\nEach mark describes an implementation and links to an interface or its documentation. This is a technical site inventory, not external certification or a promise of inclusion in AI answers.\n\n[JSON checklist]() · [Build inventory and hashes]() · [Contracts and examples]()"},{"heading":"01 / Discovery","level":3,"url":"https://lucioduran.com/#cap-discovery","content":"- Implemented: D01\n\n **Public crawling**\n\n robots.txt allows public content and advertises the sitemap. [robots.txt]()\n- Implemented: D02\n\n **XML publication map**\n\n The sitemap is built from the published catalog, including individual records. [Sitemap]()\n- Implemented: D03\n\n **Index without JavaScript**\n\n Every page has an HTML link in an index organized by content kind. [HTML]()\n- Implemented: D04\n\n **Paginated JSON catalog**\n\n Canonical URLs, kinds, languages, available dates and representation links. [HTTP / JSON]()\n- Implemented: D05\n\n **llms.txt guide**\n\n Index for clients using this proposal. Generated from the same catalog. [llms.txt]()\n- Implemented: D06\n\n **Discovery on every page**\n\n HTML and Link headers advertise alternate formats and the API catalog. [HTML / Link]()"},{"heading":"02 / Content and representations","level":3,"url":"https://lucioduran.com/#cap-representations","content":"- Implemented: F01\n\n **Server-delivered content**\n\n Main text is in the initial HTML response, before JavaScript runs. [HTML]()\n- Implemented: F02\n\n **Markdown by URL**\n\n Every page supports .md. The English root uses /index.md. [Markdown]()\n- Implemented: F03\n\n **JSON by URL**\n\n Content, headings, sections, links, language and structured data in one document. [HTTP / JSON]()\n- Implemented: F04\n\n **HTTP content negotiation**\n\n Accept selects explicit Markdown or JSON; Vary: Accept separates representations. [Accept / Vary]()\n- Implemented: F05\n\n **One source for three formats**\n\n Markdown and JSON derive from HTML main, without a separate biography for bots. [HTTP / JSON]()\n- Implemented: F06\n\n **Structure preserved**\n\n Lists, tables, code, links and image references survive serialization. [Markdown]()\n- Implemented: F07\n\n **Decoration excluded from content**\n\n Navigation, controls and decorative seals are excluded from Markdown and JSON. [HTTP / JSON]()"},{"heading":"03 / Identity, attribution and citations","level":3,"url":"https://lucioduran.com/#cap-citations","content":"- Implemented: C01\n\n **Canonical document URLs**\n\n HTML, Markdown, JSON and search results refer to the same publication. [HTTP / JSON]()\n- Implemented: C02\n\n **Anchored sections**\n\n Headings have identifiers for linking a specific passage. [HTTP / JSON]()\n- Implemented: C03\n\n **Consistent identity**\n\n Person uses a stable identifier and links to the author’s public profiles. [JSON-LD]()\n- Implemented: C04\n\n **Content-specific structured data**\n\n Profile, articles, projects and credentials publish structured data alongside visible text. [JSON-LD]()\n- Implemented: C05\n\n **Declared author and dates**\n\n Articles identify the author and preserve available publication and modification dates. [JSON Feed]()\n- Implemented: C06\n\n **Project evidence and limitations**\n\n Projects link repositories, revisions and published results with their scope. [Fuentes / Sources]()"},{"heading":"04 / Retrieval and context","level":3,"url":"https://lucioduran.com/#cap-retrieval","content":"- Implemented: R01\n\n **Search over published text**\n\n BM25 searches passages extracted from build HTML. Scores measure lexical relevance. [HTTP / JSON]()\n- Implemented: R02\n\n **Explicit filters and pagination**\n\n Language, kind, offset, limit and next support bounded, repeatable traversal. [HTTP / JSON]()\n- Implemented: R03\n\n **Results with citations and versions**\n\n Every result includes an excerpt, citationUrl and the indexed document’s contentHash. [HTTP / JSON]()\n- Implemented: R04\n\n **Read individual sections**\n\n section selects a heading and its subsections, with its own hash and source hash. [HTTP / JSON]()\n- Implemented: R05\n\n **MCP context windows**\n\n Reading supports limits and continuation by Unicode characters. The hash identifies the complete source. [MCP]()"},{"heading":"05 / Contracts and interfaces","level":3,"url":"https://lucioduran.com/#cap-interfaces","content":"- Implemented: I01\n\n **RFC 9727 API catalog**\n\n A Linkset advertises endpoints and links their contracts and documentation. [Linkset]()\n- Implemented: I02\n\n **OpenAPI 3.1 contract**\n\n Operations, parameters, limits, responses and errors are documented for HTTP clients. [OpenAPI]()\n- Implemented: I03\n\n **Document and passage JSON Schema**\n\n The two response types have distinct schemas and a describedby header. [JSON Schema]()\n- Implemented: I04\n\n **Read-only MCP**\n\n Tools to list, search and read using the official SDK and HTTP transport without persistent sessions. [MCP / HTTP]()\n- Implemented: I05\n\n **RFC 9457 structured errors**\n\n Document, search, catalog, inventory and feed APIs return problem+json for handled errors. [Problem Details]()"},{"heading":"06 / Freshness and integrity","level":3,"url":"https://lucioduran.com/#cap-freshness","content":"- Implemented: V01\n\n **SHA-256 content identity**\n\n Documents explain which fields are hashed. It identifies a version, not the truth of its claims. [HTTP / JSON]()\n- Implemented: V02\n\n **Build inventory**\n\n Paginated URLs and hashes, collection version and generation time for comparing snapshots. [HTTP / JSON]()\n- Implemented: V03\n\n **Conditional reads**\n\n ETag and If-None-Match allow 304 responses without retransmitting unchanged documents. [ETag / 304]()\n- Implemented: V04\n\n **Metadata without a response body**\n\n HEAD is available on documents, search, catalog, inventory and feeds. [HEAD]()\n- Implemented: V05\n\n **Publication feeds**\n\n JSON Feed and Atom include article text; RSS publishes summaries. Dates and links come from source files. [JSON Feed / Atom / RSS]()"},{"heading":"07 / Reading and access","level":3,"url":"https://lucioduran.com/#cap-accessibility","content":"- Implemented: A01\n\n **Explicit languages**\n\n HTML declares the content language; JSON and document responses preserve it. [HTTP / JSON]()\n- Implemented: A02\n\n **Linked EN and ES versions**\n\n hreflang links available versions and declares x-default. [HTML head]()\n- Implemented: A03\n\n **Native HTML semantics**\n\n main, headings, lists, tables and links express the content structure. [HTML]()\n- Implemented: A04\n\n **Photo references, decorative seals**\n\n Photos are retained. Project seals are empty elements with aria-hidden. [HTML / aria-hidden]()\n- Implemented: A05\n\n **Public APIs without sessions**\n\n HTTP read APIs support CORS and OPTIONS. No account, cookies or key required. [CORS / OPTIONS]()"},{"heading":"08 / Consistency and boundaries","level":3,"url":"https://lucioduran.com/#cap-boundaries","content":"- Implemented: B01\n\n **Versioned profile**\n\n Pages and APIs use the same versioned professional profile snapshot. [HTTP / JSON]()\n- Implemented: B02\n\n **Index generated from HTML**\n\n The build rejects missing or duplicate heading anchors and expected pages absent from the index. [Build]()\n- Implemented: B03\n\n **Site-scoped reading**\n\n The reader accepts known public paths and bounds source size, time and redirects. [Contrato / Contract]()\n- Implemented: B04\n\n **Content treated as source material**\n\n MCP tools identify document text as untrusted material for agent instructions. [MCP instructions]()\n- Implemented: B05\n\n **Machine-readable checklist**\n\n This list and its JSON use the same registry. Each control has an identifier and evidence link. [HTTP / JSON]()"},{"heading":"Standards, proposals and scope","level":3,"url":"https://lucioduran.com/#agent-readiness-scope","content":"HTTP, HTML, JSON Schema, OpenAPI, feeds and MCP solve specific tasks. llms.txt is a discovery proposal. Custom manifests are identified as such. WebMCP remains a developing specification and this site does not claim compatibility with it. Future capabilities will be added with a verifiable contract once they exist.\n\n[Google AI features]() · [OpenAI crawlers]() · [MCP]() · [llms.txt]() · [WebMCP]()"}],"links":[{"text":"Citable Publication","url":"https://lucioduran.com/projects/citable-publication"},{"text":"Read the project","url":"https://lucioduran.com/projects/citable-publication"},{"text":"Repository","url":"https://github.com/lucioduran/citable-publication"},{"text":"RAG Evaluation Workbench","url":"https://lucioduran.com/projects/rag-evaluation-workbench"},{"text":"Read the project","url":"https://lucioduran.com/projects/rag-evaluation-workbench"},{"text":"Repository","url":"https://github.com/lucioduran/rag-evaluation-workbench"},{"text":"Agent Replay Runtime","url":"https://lucioduran.com/projects/agent-replay-runtime"},{"text":"Read the project","url":"https://lucioduran.com/projects/agent-replay-runtime"},{"text":"Repository","url":"https://github.com/lucioduran/agent-replay-runtime"},{"text":"Agent Injection Evals","url":"https://lucioduran.com/projects/agent-injection-evals"},{"text":"Read the project","url":"https://lucioduran.com/projects/agent-injection-evals"},{"text":"Repository","url":"https://github.com/lucioduran/agent-injection-evals"},{"text":"MCP Contract Harness","url":"https://lucioduran.com/projects/mcp-contract-harness"},{"text":"Read the project","url":"https://lucioduran.com/projects/mcp-contract-harness"},{"text":"Repository","url":"https://github.com/lucioduran/mcp-contract-harness"},{"text":"All projects","url":"https://lucioduran.com/projects"},{"text":"Experience","url":"https://lucioduran.com/work-experience"},{"text":"Expertise","url":"https://lucioduran.com/areas-of-expertise"},{"text":"Education","url":"https://lucioduran.com/education"},{"text":"Biography","url":"https://lucioduran.com/bio"},{"text":"Blog","url":"https://lucioduran.com/blog"},{"text":"Projects","url":"https://lucioduran.com/projects"},{"text":"Architecture decisions","url":"https://lucioduran.com/decisions"},{"text":"CV","url":"https://lucioduran.com/cv"},{"text":"Contact","url":"https://lucioduran.com/contact"},{"text":"Complete index","url":"https://lucioduran.com/site-index"},{"text":"Content access","url":"https://lucioduran.com/access"},{"text":"Privacy","url":"https://lucioduran.com/privacy"},{"text":"email@lucioduran.com","url":"mailto:email@lucioduran.com"},{"text":"Github","url":"https://github.com/lucioduran"},{"text":"Facebook","url":"https://facebook.com/lucioduran"},{"text":"LinkedIn","url":"https://uy.linkedin.com/in/lucioduran"},{"text":"Twitter","url":"https://twitter.com/lucioduran"},{"text":"Instagram","url":"https://instagram.com/lucioduran"},{"text":"JSON checklist","url":"https://lucioduran.com/api/capabilities?locale=en"},{"text":"Build inventory and hashes","url":"https://lucioduran.com/api/publication?locale=en"},{"text":"Contracts and examples","url":"https://lucioduran.com/access"},{"text":"robots.txt","url":"https://lucioduran.com/robots.txt"},{"text":"Sitemap","url":"https://lucioduran.com/sitemap.xml"},{"text":"HTML","url":"https://lucioduran.com/site-index"},{"text":"HTTP / JSON","url":"https://lucioduran.com/api/content-index?locale=en&limit=5"},{"text":"llms.txt","url":"https://lucioduran.com/llms.txt"},{"text":"HTML / Link","url":"https://lucioduran.com/access#discovery"},{"text":"HTML","url":"https://lucioduran.com/education"},{"text":"Markdown","url":"https://lucioduran.com/education.md"},{"text":"HTTP / JSON","url":"https://lucioduran.com/education.json"},{"text":"Accept / Vary","url":"https://lucioduran.com/access#formats"},{"text":"HTTP / JSON","url":"https://lucioduran.com/bio.json"},{"text":"Markdown","url":"https://lucioduran.com/blog/rag-pipeline-architecture-chunking-reranking-evaluation.md"},{"text":"HTTP / JSON","url":"https://lucioduran.com/projects.json"},{"text":"HTTP / JSON","url":"https://lucioduran.com/projects/citable-publication.json"},{"text":"HTTP / JSON","url":"https://lucioduran.com/access.json"},{"text":"JSON-LD","url":"https://lucioduran.com/index.json"},{"text":"JSON-LD","url":"https://lucioduran.com/projects/citable-publication.json"},{"text":"JSON Feed","url":"https://lucioduran.com/feed.json"},{"text":"Fuentes / Sources","url":"https://lucioduran.com/projects/mcp-contract-harness"},{"text":"HTTP / JSON","url":"https://lucioduran.com/api/search?q=recall&locale=en&limit=3"},{"text":"HTTP / JSON","url":"https://lucioduran.com/api/search?q=agent&locale=en&kind=blog&limit=2"},{"text":"HTTP / JSON","url":"https://lucioduran.com/api/search?q=recall&locale=en&limit=1"},{"text":"HTTP / JSON","url":"https://lucioduran.com/api/content?path=/access&format=json§ion=search"},{"text":"MCP","url":"https://lucioduran.com/access#mcp"},{"text":"Linkset","url":"https://lucioduran.com/.well-known/api-catalog"},{"text":"OpenAPI","url":"https://lucioduran.com/.well-known/openapi.json"},{"text":"JSON Schema","url":"https://lucioduran.com/.well-known/schemas/publication.json"},{"text":"MCP / HTTP","url":"https://lucioduran.com/access#mcp"},{"text":"Problem Details","url":"https://lucioduran.com/.well-known/openapi.json"},{"text":"HTTP / JSON","url":"https://lucioduran.com/api/publication?locale=en&limit=3"},{"text":"ETag / 304","url":"https://lucioduran.com/access#integrity"},{"text":"HEAD","url":"https://lucioduran.com/access#integrity"},{"text":"JSON Feed / Atom / RSS","url":"https://lucioduran.com/feed.json"},{"text":"HTML head","url":"https://lucioduran.com/blog"},{"text":"HTML / aria-hidden","url":"https://lucioduran.com/projects"},{"text":"CORS / OPTIONS","url":"https://lucioduran.com/api/content-index?locale=en&limit=1"},{"text":"HTTP / JSON","url":"https://lucioduran.com/api/ai-profile"},{"text":"Build","url":"https://lucioduran.com/access#measurement"},{"text":"Contrato / Contract","url":"https://lucioduran.com/access#boundaries"},{"text":"MCP instructions","url":"https://lucioduran.com/access#mcp"},{"text":"HTTP / JSON","url":"https://lucioduran.com/api/capabilities?locale=en"},{"text":"Google AI features","url":"https://developers.google.com/search/docs/appearance/ai-features"},{"text":"OpenAI crawlers","url":"https://developers.openai.com/api/docs/bots"},{"text":"MCP","url":"https://modelcontextprotocol.io/specification/latest"},{"text":"llms.txt","url":"https://llmstxt.org/"},{"text":"WebMCP","url":"https://webmachinelearning.github.io/webmcp/"}],"structuredData":[{"@context":"https://schema.org","@graph":[{"@type":"WebSite","@id":"https://lucioduran.com/#website","url":"https://lucioduran.com","name":"Lucio Durán","publisher":{"@id":"https://lucioduran.com/#person"}},{"@type":"ProfilePage","@id":"https://lucioduran.com/#page","url":"https://lucioduran.com/","name":"Lucio Duran","inLanguage":"en","mainEntity":{"@id":"https://lucioduran.com/#person"}},{"@type":"Person","@id":"https://lucioduran.com/#person","name":"Lucio Duran","url":"https://lucioduran.com","jobTitle":"Chief Technology Officer","description":"22+ years architecting and shipping software products across healthcare, e-commerce, and marketplace verticals, from system design to production deployment.\nI lead engineering organizations where AI is not a feature but the core infrastructure layer. My work spans agentic AI systems, LLM integration patterns, full-stack development (React, React Native, Node, Python, iOS), and the strategic decisions that turn technical capabilities into revenue.\nI've built and scaled cross-functional teams across LATAM, the US, and Europe, owning the full product lifecycle: architecture decisions, CI/CD pipelines, cloud infrastructure, API design, and frontend performance. I operate at the intersection of hands-on engineering and executive technology strategy, I still read the PRs, and I still make the technical calls.\nCurrent focus: AI-augmented product development, agentic workflows, and building technology that compounds.","email":"email@lucioduran.com","sameAs":["https://github.com/lucioduran","https://facebook.com/lucioduran","https://uy.linkedin.com/in/lucioduran","https://twitter.com/lucioduran","https://instagram.com/lucioduran"],"worksFor":{"@type":"Organization","name":"LOi"}}]}],"contentHash":"4f354a9fd61688fe91aadd0b32ed085d200601beb83c8136109202c4d32bddc6","provenance":{"method":"Deterministic serialization of the server-rendered main element","source":"https://lucioduran.com/","hashAlgorithm":"sha256","hashScope":"JSON document fields before contentHash and provenance","sourceHTMLBytes":64178,"contentBytes":14894}}