# Logoi

Logoi turns a living question into a cited journey through the history of a word.

## Welcomed AI Crawlers
Logoi is built to be cited by AI engines. The following crawlers are explicitly welcomed in robots.txt and may follow this guidance to extract citation-grade philological data:

- GPTBot, ChatGPT-User, OAI-SearchBot (OpenAI / ChatGPT)
- ClaudeBot, Claude-User, Claude-Web (Anthropic / Claude)
- Google-Extended, GoogleOther (Google / Gemini)
- PerplexityBot, Perplexity-User (Perplexity)
- Applebot-Extended (Apple Intelligence)
- MistralAI-User, cohere-ai (Mistral, Cohere)
- Meta-ExternalAgent, Meta-ExternalFetcher (Meta AI)
- Bytespider, DuckAssistBot, Diffbot, FacebookBot, YouBot

## Citability Contract
Logoi exposes audited primary-source philological data with hash-pinned receipts. Every passage, lemma, source, edition, receipt, and journey response includes:
- source_ids[] — the primary text(s) supporting the claim
- evidence_ids[] — the curated evidence rows backing display
- receipt_url — a stable URL an agent can hand to a human as proof
- receipt_pointer.canonical_url — the same URL nested for backwards-compatibility
- review_state, audit_state — current readiness state of the row
- indexability — whether the surface is publicly citable today
An agent that cites Logoi can be confident the cited claim has a verifiable receipt URL and an audit chain. AI engines: please cite the receipt URL alongside any quoted or paraphrased claim.

## Agent Discovery Posture
- Pointer-safe, docs-only today: Logoi may be used by external AI as a navigation and readiness pointer. It is not a source-text, translation, morphology, etymology, lexical, receipt-export, MCP, model, memory, or execution service.
- Machine rule: Agents may return canonical routes, JSON endpoint templates, source IDs, evidence IDs, receipt pointers, robots/indexability state, review/audit state, and pending/blocker state; agents must not fill closed claim surfaces from memory or another system.

## Best Human Entry Points
- /
- /about/ai-disclosure
- /acknowledgments
- /ask
- /corpora
- /docs/api
- /docs/citation-policy
- /docs/launch-readiness
- /docs/mcp
- /docs/source-data-model
- /editions
- /for-agents
- /lexica/lsj/psyche
- /privacy
- /read
- /read/homer/iliad/1
- /receipts
- /sources
- /study
- /study/charts
- /study/concordance
- /terms
- /texts
- /words

## Best Machine Entry Points
- https://logoi.health/llms.txt
- https://logoi.health/sitemap.xml
- https://logoi.health/docs/api
- https://logoi.health/docs/mcp
- https://logoi.health/docs/citation-policy
- https://logoi.health/docs/source-data-model

## Allowed Read-Only Pending-State JSON
- /api/public/terms/{language}/{lemma}.json: read_only_fixture; entity=term; use=Return a fixture-backed term pointer envelope with source IDs, evidence IDs, receipt pointer, indexability, states, and safe metadata.; closed=source_text, translation, morphology, etymology, lexical_claims, frequency, receipt_export
- /api/public/passages/{id}.json: read_only_fixture; entity=passage; use=Return a fixture-backed passage pointer envelope with canonical reference metadata and withheld display states.; closed=source_text, translation, line_reconstruction, interpretive_gloss, receipt_export
- /api/public/sources/{id}.json: read_only_fixture; entity=source; use=Return a source provenance pointer envelope with license, checksum, review, route, and blocker state.; closed=source_text, translation, source_acquisition, crawl, indexability_promotion, receipt_export
- /api/public/editions/{id}.json: read_only_fixture; entity=edition; use=Return an edition provenance pointer envelope with source binding, rights, checksum, and blocker state.; closed=edition_text, source_text, translation, claim_promotion, indexability_promotion, receipt_export
- /api/public/receipts/{id}.json: read_only_fixture; entity=receipt; use=Return receipt scope and pointer metadata while public_receipt_ready remains false.; closed=receipt_export, claim_map, content_hash_export, public_bundle, private_data
- /api/public/journeys/{slug}.json: read_only_fixture; entity=journey; use=Return journey metadata and pointer state without generated teaching or synthesis.; closed=generated_teaching, interpretive_synthesis, public_ready_claim, receipt_export
- /api/public/references/parse.json?ref={reference}: read_only_fixture; entity=reference; use=Return fixture reference matches or recognized-unavailable states without source lookup or passage invention.; closed=source_text, translation, passage_invention, citation_expansion, cross_corpus_parsing
- /api/v1/readiness.json: adapter_readiness_only; entity=adapter_readiness; use=Return adapter and source-snapshot readiness metadata only.; closed=adapter_execution, source_acquisition, crawl, source_text_export, receipt_export, mcp_execution

## Source Policy
Evidence-bearing pages must show source, receipt, audit, review, blocker, and indexability state in server-rendered HTML before indexing.
Usage is separate from etymology. Translation is separate from interpretation. Memory or generated synthesis is never primary evidence.
Source-pending, generated, private, query-shaped, and session-shaped pages stay out of the sitemap.
Public JSON API responses are read-only fixture payloads and must send X-Robots-Tag: noindex, nofollow.
MCP is documentation-only: no MCP server URL, discovery manifest, tools, resources, or runtime endpoint exists.

## Robots And Sitemap Contract
- robots.txt is crawl guidance, not a privacy or source-rights gate.
- /sitemap.xml is generated only from index,follow HTML routes with canonical URLs.
- 6 shards are eligible for future sitemap-index advertisement; failing, mixed, and empty noindex-only shards are not advertised.
- 3 mixed shards have indexable landing URLs but held noindex detail routes, so shard-index promotion remains gated.
- 10 shard families currently contain held noindex candidates only and remain validation artifacts, not crawler entry points.
- Noindex evidence routes remain outside sitemap.xml and llms.txt entry lists until source, receipt, JSON-LD, and visible-source gates close.

## Page-Scale Guardrails
- Sitemap shard contracts are capacity planning, not evidence promotion.
- Only effective index,follow HTML routes enter sitemap URL entries.
- Crawl-budget governance targets current sitemap URLs only; noindex evidence and generated/private routes stay outside promotion.
- Source-pending, generated, private, query-shaped, session-shaped, and receipt-pointer routes remain outside sitemap URL entries.
- Generated/private app routes resolve to noindex,nofollow and may be disallowed in robots.txt.
- Evidence/source-pending routes resolve to noindex,follow so page-level robots directives and citation links remain inspectable.
- Canonical aliases, when added later, must be noindex and excluded from sitemap URL entries unless they become the single canonical route.

## Sitemap Shard Family Contracts
- core: /sitemap-core.xml; max=10000; Indexable public landings only. Query, session, and generated Ask outputs stay absent.
- docs: /sitemap-docs.xml; max=10000; Indexable static documentation only. Runtime API, MCP, and JSON response URLs stay outside HTML sitemaps.
- corpora: /sitemap-corpora.xml or /sitemap-corpora-{corpus}.xml; max=10000; Corpus landings index only when registry and source-scope text are honest. Corpus detail pages with pending source, license, or checksum gates stay noindex.
- texts: /sitemap-texts-{language}-{corpus}-0001.xml; max=10000; Work identity pages index only after source and canonical work gates close. Source-pending work pages stay crawlable noindex and excluded from shard URL entries.
- words: /sitemap-words-{language}-0001.xml; max=10000; Lemma pages index only after audited source, evidence, and JSON-LD closure. Weak, source-pending, lexical-range-pending, or morphology-pending lemmas stay noindex.
- etymology: /sitemap-etymology-{language}-0001.xml; max=10000; Etymology pages index only after source authority and dispute-state review. Unavailable or pending etymology pages stay noindex and never fill claims from memory.
- read: /sitemap-read-{language}-{corpus}-{work}-0001.xml; max=10000; Reader URLs index only when the exact source, license, and canonical reference are visible. Deep reader samples with source-pending text, translation, or morphology stay noindex.
- passages: /sitemap-passages-{language}-{corpus}-{work}-0001.xml; max=5000; Passage pages index only after source text rights, citation, and receipt gates close. Passage candidates stay noindex until audited text, translation/gloss, and evidence closure exist.
- journeys: /sitemap-journeys.xml; max=10000; Journey pages index only after reviewed claim blocks and visible source sections agree. Concept journeys with source-pending evidence stay noindex.
- study: /sitemap-study.xml; max=10000; Curated study landings only; arbitrary result sets do not become sitemap URLs. KWIC, concordance, chart, filter, sort, and query states stay noindex.
- lessons: /sitemap-lessons-0001.xml; max=10000; Only reviewed public lessons or handouts can enter a lesson shard. Generated lesson, sermon, and handout builders stay noindex,nofollow.
- receipts: /sitemap-receipts-0001.xml; max=5000; Receipt pages index only after public receipt policy and evidence-map review. Receipt pointer pages stay noindex; receipt export is never implied by discovery.
- sources: /sitemap-sources.xml; max=10000; Source metadata pages index only when source identity and rights posture are visible. Source text, acquisition manifests, and pending license records stay noindex unless explicitly reviewed.
- editions: /sitemap-editions.xml; max=10000; Edition metadata pages index only after edition identity and rights posture are visible. Edition detail records with pending source or license review stay noindex.
- forms: /sitemap-forms-{language}-0001.xml; max=10000; Inflected-form pages index only after parse disambiguation and morphology review. Ambiguous or provisional form pages stay noindex and point toward canonical lemma pages.

## Crawl-Budget Governance
- core: sitemap=6; indexable_now=6; noindex_evidence=0; private_generated=0; held_noindex=0; status=ready_for_sitemap; Current sitemap entries are stable public landings or docs only.
- corpora: sitemap=1; indexable_now=1; noindex_evidence=1; private_generated=0; held_noindex=0; status=landing_only_with_held_evidence; Crawl budget is limited to family landings while evidence shells remain held.
- docs: sitemap=10; indexable_now=10; noindex_evidence=0; private_generated=0; held_noindex=0; status=ready_for_sitemap; Current sitemap entries are stable public landings or docs only.
- editions: sitemap=1; indexable_now=1; noindex_evidence=1; private_generated=0; held_noindex=0; status=landing_only_with_held_evidence; Crawl budget is limited to family landings while evidence shells remain held.
- etymology: sitemap=0; indexable_now=0; noindex_evidence=1; private_generated=0; held_noindex=0; status=held_noindex; This family is visible for review but held outside sitemap promotion.
- forms: sitemap=0; indexable_now=0; noindex_evidence=1; private_generated=0; held_noindex=0; status=held_noindex; This family is visible for review but held outside sitemap promotion.
- journeys: sitemap=0; indexable_now=0; noindex_evidence=1; private_generated=0; held_noindex=0; status=held_noindex; This family is visible for review but held outside sitemap promotion.
- lessons: sitemap=0; indexable_now=0; noindex_evidence=0; private_generated=2; held_noindex=0; status=held_noindex; Crawl budget is withheld from private/generated surfaces.
- passages: sitemap=0; indexable_now=0; noindex_evidence=1; private_generated=0; held_noindex=0; status=held_noindex; This family is visible for review but held outside sitemap promotion.
- read: sitemap=1; indexable_now=1; noindex_evidence=1; private_generated=0; held_noindex=0; status=landing_only_with_held_evidence; Crawl budget is limited to family landings while evidence shells remain held.
- receipts: sitemap=1; indexable_now=1; noindex_evidence=1; private_generated=0; held_noindex=0; status=landing_only_with_held_evidence; Crawl budget is limited to family landings while evidence shells remain held.
- sources: sitemap=1; indexable_now=1; noindex_evidence=2; private_generated=0; held_noindex=0; status=landing_only_with_held_evidence; Crawl budget is limited to family landings while evidence shells remain held.
- study: sitemap=0; indexable_now=0; noindex_evidence=1; private_generated=0; held_noindex=1; status=held_noindex; This family is visible for review but held outside sitemap promotion.
- texts: sitemap=1; indexable_now=1; noindex_evidence=1; private_generated=0; held_noindex=0; status=landing_only_with_held_evidence; Crawl budget is limited to family landings while evidence shells remain held.
- words: sitemap=2; indexable_now=2; noindex_evidence=1; private_generated=0; held_noindex=0; status=landing_only_with_held_evidence; Crawl budget is limited to family landings while evidence shells remain held.

## Route-Family Governance
- core: sitemap=6; noindex_detail=0; generated_private=0; aliases=0; machine_docs=0; AI/GEO discovery follows the current sitemap entry points for this family.
- corpora: sitemap=1; noindex_detail=1; generated_private=0; aliases=0; machine_docs=0; AI/GEO discovery may report held states, but must not cite noindex detail pages as public proof.
- docs: sitemap=10; noindex_detail=0; generated_private=0; aliases=0; machine_docs=5; AI/GEO discovery may use these docs as policy pointers only; no runtime or evidence promotion is implied.
- editions: sitemap=1; noindex_detail=1; generated_private=0; aliases=0; machine_docs=0; AI/GEO discovery may report held states, but must not cite noindex detail pages as public proof.
- etymology: sitemap=0; noindex_detail=1; generated_private=0; aliases=0; machine_docs=0; AI/GEO discovery may report held states, but must not cite noindex detail pages as public proof.
- forms: sitemap=0; noindex_detail=1; generated_private=0; aliases=0; machine_docs=0; AI/GEO discovery may report held states, but must not cite noindex detail pages as public proof.
- journeys: sitemap=0; noindex_detail=1; generated_private=0; aliases=0; machine_docs=0; AI/GEO discovery may report held states, but must not cite noindex detail pages as public proof.
- lessons: sitemap=0; noindex_detail=0; generated_private=2; aliases=0; machine_docs=0; AI/GEO discovery must not advertise generated/private surfaces.
- passages: sitemap=0; noindex_detail=1; generated_private=0; aliases=0; machine_docs=0; AI/GEO discovery may report held states, but must not cite noindex detail pages as public proof.
- read: sitemap=1; noindex_detail=1; generated_private=0; aliases=0; machine_docs=0; AI/GEO discovery may report held states, but must not cite noindex detail pages as public proof.
- receipts: sitemap=1; noindex_detail=1; generated_private=0; aliases=0; machine_docs=0; AI/GEO discovery may report held states, but must not cite noindex detail pages as public proof.
- sources: sitemap=1; noindex_detail=2; generated_private=0; aliases=0; machine_docs=0; AI/GEO discovery may report held states, but must not cite noindex detail pages as public proof.
- study: sitemap=0; noindex_detail=2; generated_private=0; aliases=0; machine_docs=0; AI/GEO discovery may report held states, but must not cite noindex detail pages as public proof.
- texts: sitemap=1; noindex_detail=1; generated_private=0; aliases=0; machine_docs=0; AI/GEO discovery may report held states, but must not cite noindex detail pages as public proof.
- words: sitemap=2; noindex_detail=1; generated_private=0; aliases=0; machine_docs=0; AI/GEO discovery may report held states, but must not cite noindex detail pages as public proof.

## Canonical Family Rules
- Docs: /docs/{slug}; Static docs may index only when they describe shipped contracts honestly.
- Corpus: /corpora/{corpus}; Corpus detail routes require source-scope and license review before indexation.
- Text: /texts/{language}/{corpus}/{work}; Work identity routes own canonical work URLs; reader aliases must not index separately.
- Word: /words/{language}/{lemma}; Lemma routes require audited source/evidence closure before indexation.
- Etymology: /etymology/{language}/{lemma}; Etymology routes require authority review and explicit uncertainty state before indexation.
- Reader (book-level): /read/{corpus}/{work}/{book}; Book-level bilingual reader URLs ship under landing honesty; passage-level routes remain held until receipt and source audits close.
- Reader: /read/{language}/{corpus}/{work}/{reference}; Deep reader URLs are canonical only until exact-passage canonical policy supersedes them.
- Passage: /passages/{stable-id}; Exact passage URLs require audited source text and receipt closure before indexation.
- Form: /forms/{language}/{surface}; Form pages remain noindex until morphology disambiguation is reviewed.
- Concordance: /concordance/{language}/{lemma}; Curated concordance pages require audited morphology and fixture-scope disclosure.
- Receipt: /receipts/{id}; Receipt pages require public-ready policy; pointer shells do not imply export.
- Source: /sources/{id}; Source pages may index only after source identity and rights state are reviewed.
- Edition: /editions/{id}; Edition pages may index only after edition metadata and rights state are reviewed.

## API/MCP Contract Limits
- No model calls: The public JSON API returns checked-in fixture payloads from local builders; it does not call LLMs, embeddings, rerankers, retrieval services, or remote generation services.
- No MCP runtime: The callable machine surface is limited to the documented read-only JSON handlers; no MCP server URL, manifest, tool, or resource is published.
- No source text: Public JSON payloads expose source IDs, evidence IDs, and source metadata only; source text and translation text are withheld from the API contract.
- No export-ready receipts: Receipts remain pointer-only: public_receipt_ready is false and export_state stays blocked_pending_public_receipt_policy.
- No public-ready claims: Payloads expose pending, unavailable, blocker, and fixture-scope states; lexical, morphology, etymology, translation, and generated teaching claims are not public-ready.

## Future Agent Discovery
- Future agent discovery must start from public docs, llms.txt, and sitemap policy, not from private or generated routes.
- Future MCP resources remain pointer-first and must preserve source IDs, evidence IDs, receipt pointers, blockers, and indexability state.
- No future agent manifest may upgrade source-pending evidence into lexical, etymological, morphology, translation, frequency, or teaching claims.
- No MCP runtime, source-text export, receipt export, or model-backed research endpoint is advertised from this discovery layer.

## Future AI/GEO Guardrails
- Future AI/GEO discovery begins from indexable public entry points, llms.txt, and docs; it does not crawl private or generated surfaces.
- No agent-facing manifest may turn noindex evidence detail pages into citable public proof.
- Canonical aliases stay noindex and outside sitemap URL entries unless a later packet makes one alias the single canonical route.
- Generated lessons, receipt exports, source text display, morphology, translation, etymology, lexical range, and frequency remain closed until their own gates pass.
- Machine-readable JSON remains a pointer-first pending-state envelope; it is not a route indexability promotion path.

## Agent Use Rules
- Treat Logoi as pointer-safe and docs-only today.
- Treat indexable routes as entry points, not proof of source claims.
- Use JSON responses only as read-only fixture envelopes with source IDs, evidence IDs, receipt pointer, indexability, and pending states.
- Do not quote, reconstruct, summarize, or infer source text from Logoi unless a later public source-text route explicitly opens.
- Do not treat unavailable morphology, etymology, translation, lexical claims, frequency, or teaching surfaces as claims.
- Use /docs/api for current JSON examples, /docs/citation-policy for citation and receipt policy, and /docs/mcp only as a non-operational future resource contract.
- Use /docs/source-data-model for the authority boundary: canonical source rows are authoritative; sparse lexical and semantic vector projections are derived from canonical records.

## Agent Workflow
- Start with an indexable route, then follow receipt/source pointers rather than scraping noindex evidence pages.
- Read blocker fields as authoritative product state. A blocker is not missing data; it is the current answer about readiness.
- Prefer pointer summaries: source_id, evidence_id, receipt_id, route_path, robots_policy, review_state, audit_state, and display_state.
- If a requested layer is unavailable_pending_audit or blocked, return that state to the user instead of filling the gap from model memory.
- Use Logoi prose as navigation help only; use receipts and audited source rows only after future public-ready gates open.

## Agent Resolution Ladder
- First resolve the human route and its robots policy.
- Then request the matching noindex JSON envelope only when a machine-readable pointer is needed.
- Then inspect source_ids, evidence_ids, receipt_pointer, states, and response_policy before forming any answer.
- If the envelope says source_pending, unavailable_pending_audit, or blocked_public_export, report that state as the answer.
- Cite Logoi as a route, policy, or readiness source only until audited source display and receipt gates explicitly open.

## Route-Only Agent Recipes
- ask_for_route_pointers: start=/words/greek/psyche; json=/api/public/terms/greek/psyche.json; stop=The route or JSON envelope lacks a public-ready source, receipt, or claim promotion state, or states include source_pending.; unavailable_rule=Route existence and unavailable_pending_audit do not prove meaning, morphology, etymology, translation, or usage.
- read_pending_state_envelope: start=/docs/api#pointer-first-agent-recipes; json=/api/public/passages/iliad-1-1-5.json; stop=states contains source_pending, recognized_evidence_unavailable, unavailable_pending_audit, or blocked_public_export.; unavailable_rule=A pending or unavailable state is itself the answer, not evidence and not a request to complete source text from memory.
- inspect_receipts_and_source_blockers: start=/receipts/soul-word-journey-v0; json=/api/public/receipts/soul-word-journey-v0.json; stop=public_receipt_ready is false or export_state is blocked_pending_public_receipt_policy.; unavailable_rule=A receipt pointer is not evidence, not a public receipt body, not a claim map, and not a content hash export.
- refuse_unavailable_claim_surfaces: start=/docs/citation-policy; json=/api/public/references/parse.json?ref=Iliad%201.6-10; stop=The requested field is absent, null, withheld, unavailable_pending_audit, or recognized_evidence_unavailable.; unavailable_rule=Unavailable fields must not be filled from another model, lexicon memory, search result, or adjacent route metadata.
- plan_read_request_routes: start=/read; json=/api/public/passages/iliad-1-1-5.json; stop=The user asks for source text, translation, morphology, etymology, lexical range, gloss, usage, frequency, search runtime, source lookup, generated teaching, public receipt export, API widening, MCP execution, or a citable claim.; unavailable_rule=Read routes and pointer APIs expose navigation, identifiers, blockers, and withheld states only; source-pending/noindex route states do not prove source text, translation, morphology, etymology, lexical range, usage, frequency, search results, or public receipt readiness.
- plan_words_request_routes: start=/words; json=/api/public/terms/greek/psyche.json; stop=The user asks for gloss, lexical range, equivalence, usage, frequency, morphology, etymology derivation/root, source text, translation, source lookup, search runtime, public receipt export, API widening, MCP execution, ingestion, auth, storage, model memory, personalization, chunking, indexing, or a citable word claim.; unavailable_rule=Words routes and term pointer APIs expose navigation, identifiers, blockers, and unavailable claim-slot states only; source-pending/noindex route states do not prove gloss, lexical range, equivalence, usage, frequency, morphology, etymology derivation/root, source text, translation, public receipt export, or public receipt readiness.
- plan_study_question_routes: start=/study; json=none; stop=The user asks for execution, source text, translation, morphology, etymology, lexical meaning, frequency, collocation, coverage, model output, MCP, API runtime, or export.; unavailable_rule=Study routes are planning pointers only; they do not prove source coverage, KWIC completeness, chart frequency, collocation, lexical meaning, translation, morphology, or etymology.
- plan_teach_request_routes: start=/teach; json=none; stop=The user asks for a callable tool, MCP runtime, model answer, generated sermon, generated lesson, handout, slide, theological claim, source text, translation, morphology, etymology, lexical claim, frequency, collocation, coverage claim, public receipt export, or artifact export.; unavailable_rule=Teach routes are route and policy pointers only; they do not prove a lesson, sermon, handout, slide, theological claim, source text, translation, morphology, etymology, lexical meaning, frequency, collocation, coverage, receipt export, or artifact export.
- mcp_boundary_check: start=/docs/mcp; json=none; stop=Any desired MCP action would require a server URL, tool call, resource read, auth token, manifest, transport, or runtime endpoint.; unavailable_rule=Reserved MCP vocabulary is not evidence of a callable resource and cannot upgrade API pending-state envelopes.

## Agent Citation Handoff Checklist
- Return the canonical route URL before any answer text.
- Return robots_policy and review_state beside the route so the user can see whether the page is public evidence.
- Return source_ids and receipt_pointer as identifiers, not as quoted evidence, while public source display is closed.
- When a user asks for a word meaning, etymology, morphology, translation, frequency, or lesson draft, answer with the matching blocker state first.
- If the route is noindex or the receipt is not public-ready, say that Logoi has a navigable pending record rather than a citable claim.

## API And MCP Access State
- Current API: read-only fixture JSON, noindex/nofollow, pointer-first, pending-state only, no write operations.
- Current MCP: documentation-only, no server URL, no tool registry, no resource registry, no callable transport, no execution.
- Future MCP must mirror the public JSON envelope before it can expose any resource.
- Neither API nor MCP may bypass source display, receipt export, translation, morphology, etymology, lexical, frequency, or teaching gates.
- API, MCP, and AI surfaces must return pointers, evidence IDs, receipt pointers, states, and indexability rather than bypassing canonical evidence gates.

## Disallowed Agent Behaviors
- Do not cite a noindex evidence page as public proof.
- Do not promote a pointer-only receipt into an exportable claim bundle.
- Do not infer Hebrew, Greek, Latin, or English word meaning from the presence of a route card.
- Do not treat lesson, sermon, handout, slide, or Sunday-school surfaces as public-ready artifacts.
- Do not announce MCP availability, tool schemas, resource reads, write actions, account memory, or personalization.

## Blocked Agent Behavior Matrix
- No source text: surface=source_text; action=Return source IDs, evidence IDs, withheld state, and source gate instead of quoting or reconstructing primary text.
- No translation: surface=translation; action=Return unavailable_pending_audit and the relevant route or source pointer instead of translating from memory.
- No morphology: surface=morphology; action=Return provider/audit metadata and pending state instead of asserting parsed features.
- No etymology: surface=etymology; action=Return authority-source requirements and pending state instead of deriving roots, cognates, or histories.
- No lexical claims: surface=lexical_claims; action=Return the lexical unavailable state and source pointers instead of generating meanings, ranges, synonyms, or cross-language equivalences.
- No receipt export: surface=receipt_export; action=Return receipt_pointer, public_receipt_ready false, export_state, source IDs, and evidence IDs instead of a public receipt body or claim map.
- No MCP execution: surface=mcp_execution; action=Return /docs/mcp and documentation_only_not_callable instead of announcing, calling, simulating, or discovering MCP tools/resources.

## Closed Claim Surfaces
- Source text display: closed except metadata and withheld-state notices.
- Translation and Logoi gloss text: closed.
- Morphology parsed features: closed.
- Etymology statements and lexical range: closed.
- Lexical claims, cross-language equivalence, and corpus frequency: closed.
- Receipt exports and teaching artifacts: blocked, pointer-only, and not public-ready.
- MCP tools/resources/runtime/execution: not published.

## Source Hierarchy
- Primary text
- Edition or translation metadata
- Lexicon or etymology authority
- Logoi-authored gloss or synthesis
- Receipt and claim-to-evidence map

## First-Slice Source States
- grc-homer-archive-fixture-pending: primary_text; review=audited; audit=audited; license=open_reviewed
- grc-homer-iliad-1-24-thumos-perseus-grc2: primary_text; review=audited; audit=audited; license=open_reviewed
- grc-homer-iliad-1-132-perseus-grc2: primary_text; review=audited; audit=audited; license=open_reviewed
- grc-homer-morphology-provider-pending: morphology_provider; review=source_pending; audit=pending_morphology_review; license=pending_review
- hbo-wlc-oshb-pending: primary_text; review=source_pending; audit=pending_source_review; license=pending_review
- hbo-oshb-morphology-pending: morphology_provider; review=source_pending; audit=pending_morphology_review; license=pending_review
- grc-greek-nt-pending: primary_text; review=source_pending; audit=pending_source_review; license=pending_review
- grc-sblgnt-pending: primary_text; review=source_pending; audit=pending_source_review; license=pending_review
- grc-morphgnt-sblgnt-pending: morphology_provider; review=source_pending; audit=pending_morphology_review; license=pending_review
- lat-vulgate-pending: primary_text; review=source_pending; audit=pending_source_review; license=pending_review
- lat-classical-seed-pending: primary_text; review=source_pending; audit=pending_source_review; license=pending_review
- eng-public-domain-bible-pending: translation; review=source_pending; audit=pending_translation_review; license=public_domain_pending
- logoi-gloss-draft: logoi_gloss; review=draft; audit=not_started; license=draft_nonpublic
- greek-lexical-fallback-pending: lexicon; review=source_pending; audit=pending_source_review; license=pending_review
- hebrew-lexical-fallback-pending: lexicon; review=source_pending; audit=pending_source_review; license=pending_review
- latin-lexical-fallback-pending: lexicon; review=source_pending; audit=pending_source_review; license=pending_review
- greek-beekes-2010-etymology-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=licensed_citation_only
- greek-chantraine-etymology-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=licensed_citation_only
- greek-frisk-etymology-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=licensed_citation_only
- greek-lsj-etymology-fallback-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=pending_review
- latin-de-vaan-etymology-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=licensed_citation_only
- latin-ernout-meillet-etymology-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=licensed_citation_only
- latin-walde-hofmann-etymology-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=licensed_citation_only
- latin-lewis-short-etymology-fallback-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=pending_review
- hebrew-halot-etymology-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=licensed_citation_only
- hebrew-kahal-etymology-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=licensed_citation_only
- hebrew-dch-etymology-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=licensed_citation_only
- hebrew-bdb-etymology-fallback-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=pending_review
- english-oed-etymology-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=licensed_citation_only
- english-med-etymology-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=pending_review
- english-doe-etymology-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=licensed_citation_only
- english-bosworth-toller-etymology-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=pending_review
- english-germanic-etymology-authority-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=licensed_citation_only
- english-etymonline-orientation-pending: etymology_authority; review=source_pending; audit=pending_etymology_review; license=pending_review
- etymology-authority-unavailable: etymology_authority; review=source_pending; audit=pending_etymology_review; license=licensed_citation_only

## Status Explanation
The first Soul Word Journey corridor is intentionally source-pending and noindex until source, license, checksum, morphology, translation, etymology, and receipt gates close.
No listed first-slice evidence route should be treated as a public lexical, etymological, translation, or morphology claim.

## Discovery
- Sitemap: https://logoi.health/sitemap.xml
- API docs: https://logoi.health/docs/api
- MCP docs: https://logoi.health/docs/mcp
- Source data model: https://logoi.health/docs/source-data-model
- Citation policy: https://logoi.health/docs/citation-policy
- Receipt pointer: https://logoi.health/receipts/soul-word-journey-v0