[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8,"_9":10},"actionData","errors","root",{"_341":342,"_343":344},"routes/_app.x.$handle.$name",{"_98":97,"_31":43,"_107":88,"_26":108,"_13":14,"_109":-5,"_11":110,"_111":112,"_113":114,"_115":116,"_117":118,"_20":-5,"_119":-5,"_120":-5,"_121":24,"_122":123,"_124":57,"_125":-5,"_126":-5,"_127":128},"routes/_app.x.$handle.$name.branch.$branch.version.$version.diff",{"_11":12,"_13":14,"_15":16,"_17":-5,"_18":-5},"project",{"_41":42,"_31":43,"_36":44,"_45":-5,"_46":24,"_47":48,"_49":48,"_50":51,"_52":-5,"_18":53,"_54":55,"_56":57,"_58":-5,"_59":-5,"_60":61,"_62":63,"_64":65,"_66":67,"_68":69,"_70":-5,"_71":72,"_73":72,"_74":75,"_76":75,"_77":75,"_78":72,"_79":72,"_80":81,"_82":83,"_84":85,"_86":24},"version",25,"changes",[19],"aiChatInfo","userId",{"_20":21,"_22":-7,"_23":24,"_25":24,"_26":27,"_28":27},"file",{"_29":30,"_31":32,"_33":20,"_34":35,"_36":37,"_38":14,"_39":40,"_23":24},"previous","deleted",false,"deletedHere","path","docs/INGESTION_SEMANTIC_RECONCILIATION_PROTOCOL.md","latestPath","fileId","34e640de-aaa0-11f1-b17a-1607ee4eb77e","name","INGESTION_SEMANTIC_RECONCILIATION_PROTOCOL.md","fileType","content","# Ingestion & Semantic Reconciliation Protocol\n\nApplies to the 11 official shallow ZIP inputs. This protocol is deterministic-first and loss-averse.\n\n## 0. Input authority\nThe outer transport container is not an input package. Exactly 11 shallow inner ZIPs are the official current input set.\n\nNo final artifact may be generated until all 11 have a physical identity and every recursively discovered manifestation has a terminal disposition.\n\n## 1. Archive safety before extraction\nFor outer and every nested archive:\n1. compute SHA-256 of archive bytes;\n2. verify archive structure/CRC where format supports it;\n3. inventory entries without extracting;\n4. reject/flag absolute paths, `..`, drive paths, ADS/colon hazards, symlink/special-device entries, Windows reserved names, trailing dot/space, duplicate paths, Unicode normalization collision and case-fold collision;\n5. enforce entry count, total uncompressed size, per-entry size and suspicious compression-ratio controls;\n6. use clean staging directory; never extract over an existing authoritative tree;\n7. hash archive before and after validation/extraction to detect archive TOCTOU.\n\nArchive failure becomes evidence and disposition; it is not silent omission.\n\n## 2. Manifestation-level ledger\nEvery physical file occurrence receives its own `manifestation_id` even when its bytes duplicate another file.\n\nRequired fields:\n`source_zip_id, nested_archive_chain, original_path, original_filename, sha256, bytes, mime_guess, extension, detected_format, encoding_state, parse_state, read_state, semantic_family, topics, lineage, exact_duplicate_group, semantic_cluster, conflict_group, canonical_object_id, destination_package, destination_path, transformation, applied_state, test_state, manual_destination, provenance, disposition`.\n\nInvariant:\n`CANONICAL_OBJECT_COUNT != SOURCE_COVERAGE`.\n\nDeduplication may reduce canonical content objects but never source manifestations or provenance paths.\n\n## 3. Actual-format detection\nDo not trust filename extension alone.\n\nUse, in order when practical:\n1. magic/header bytes;\n2. container internals;\n3. structural parser probe;\n4. extension;\n5. MIME guess.\n\nRecord `EXTENSION_FORMAT_MISMATCH` instead of silently renaming originals.\n\nOriginal bytes and source name stay preserved. Normalized derivatives get separate IDs and provenance.\n\n## 4. Encoding protocol\n`UTF8_DECODE_FAILURE != FILE_CORRUPTION`.\n\nFor textual candidates:\n1. BOM-aware UTF-8/UTF-16/UTF-32 probe;\n2. strict UTF-8;\n3. format parser on candidate decode;\n4. controlled CP1252/Latin-1 compatible fallback when evidence supports it;\n5. record detected/ambiguous encoding and confidence;\n6. binary/unreadable remains byte-read with explicit parse blocker.\n\nDo not lossy-replace undecodable bytes to manufacture a successful parse.\n\n## 5. Recursive nested-archive exhaustion\nFor each relevant archive manifestation:\n`archive -> inventory -> nested archive discovery -> safe recurse -> member ledger`.\n\nNo artificial depth cutoff. Resource controls may pause/checkpoint a branch, but do not redefine deep entries as absent.\n\nDetect repeated-content archive recursion by archive hash lineage to avoid cycles/redundant work.\n\n## 6. Deterministic analysis before LLM semantic reading\n`PHYSICAL INVENTORY -> HASH/NORMALIZED HASH -> STRUCTURAL PARSE -> EXACT DUP GROUPS -> STRUCTURAL FINGERPRINTS -> VERSION/LINEAGE SIGNALS -> WRITER/IMPORT/REFERENCE GRAPH -> STATIC DEFECT SCAN -> PRIORITY QUEUE -> SEMANTIC REASONING`.\n\nNever use an LLM for hashing, counting, joins or deterministic parsing when code can do it reliably.\n\n## 7. Read-state machine\nStates remain independent:\n`FOUND -> CENSUSED -> BYTE_READ -> PARSED -> SEMANTICALLY_READ -> CLASSIFIED -> RECONCILED -> APPLIED -> TESTED -> CURRENT`.\n\nA file may be `BLOCKED_WITH_REASON` at any stage, but never silently disappear.\n\n`FOUND != READ != UNDERSTOOD != CLASSIFIED != APPLIED != TESTED`.\n\n## 8. Duplicate model\n### EXACT_DUPLICATE\nSame SHA-256. Share canonical content if useful; retain every manifestation/path.\n\n### SEMANTIC_DUPLICATE\nDifferent bytes but materially same responsibility. Merge only after comparing purpose, scope, triggers, preconditions, authority, workflow, outputs, verification, failures, stop conditions, examples, platform semantics and version lineage.\n\n### RELATED_BUT_DISTINCT\nOverlap exists but responsibilities differ. Preserve separate identities with typed relation.\n\n### VERSIONED_SUPERSESSION\nNewer/current object supersedes older authority while older evidence remains available in provenance/history.\n\n### CONFLICT\nMaterial claims/semantics disagree. Preserve both, identify authority/currentness, and resolve explicitly or keep unresolved state.\n\nNever merge by filename, topic label or embedding proximity alone.\n\n## 9. Canonical semantic object\nMinimum fields:\n`canonical_id, canonical_name_ptbr, english_term, object_class, definition, scope, what_it_is, what_it_is_not, aliases, positive_examples, counterexamples, parents, children, related_objects, source_manifestations, platform_variants, lineage, conflicts, authority_state, currentness_state, operational_destinations`.\n\n## 10. Semantic role precedes file format\nDetermine responsibility before projecting a platform-specific file.\n\nExamples of distinct roles:\n`persistent instruction/context, Skill, Rule/Policy, Agent/Subagent, Hook, Workflow/Command, MCP config, Memory, Retriever index, Validator, Test, Template, Generator, Ledger, Receipt, Report, Status Doc`.\n\n`PLATFORM_FILE_FORMAT != SEMANTIC_ROLE`.\n\nAdapters are generated from canonical roles; canonical knowledge is not mutated to mimic one vendor filesystem.\n\n## 11. Source-to-operation closure\nFor every material reusable concept:\n`SOURCE -> CONCEPT -> WHAT IT CHANGES -> COMPONENT CLASS -> LIVE COMPONENT -> CONSUMER -> ROUTER/TRIGGER -> ENFORCEMENT -> DETECTOR -> TEST -> MUTATION -> RUNTIME EVIDENCE -> STATUS`.\n\nReverse audit:\n`ACTIVE COMPONENT -> WHY IT EXISTS -> SOURCE KNOWLEDGE -> CONTRACT -> CONSUMER -> TRIGGER -> OUTPUT -> ENFORCEMENT -> DETECTOR -> EVIDENCE`.\n\nOrphans remain REVIEW/UNKNOWN and block unrestricted completeness claims.\n\n## 12. Non-regression\nCompare current input/final packages against historical version-scoped baselines.\n\nCardinality is a signal, not a completion oracle. A smaller canonical representation may pass only when:\n- every source manifestation is dispositioned;\n- provenance remains complete;\n- material deep knowledge is local/referenced/routable;\n- routing/retrieval behavior is proven;\n- removed/superseded identities have explicit crosswalk;\n- no operational capability disappears without justified disposition.\n\n## 13. Currentness\nHistorical provider/platform syntax never proves current adapter behavior.\n\nSeparate:\n`DOC_CURRENT -> BINARY_INSTALLED -> CONNECTED -> AUTHENTICATED -> AUTHORIZED -> ENTITLED -> QUOTA_AVAILABLE -> RUNTIME_PROBED -> VERIFIED_FOR_TASK`.\n\nAdapter generation uses current official docs; runtime claims require actual probes when the runtime is available.\n\n## 14. Incremental checkpoints\nPersist after each atomic phase:\n`INPUT_CENSUS, PARSE, SEMANTIC_READ, RECONCILIATION, ARCHITECTURE, IMPLEMENTATION, TEST, PREFREEZE, FREEZE, CLEANROOM, FINAL_RECONCILIATION`.\n\nA timeout/stream interruption after a persisted phase is not task failure and does not justify repeating completed work.\n\n## 15. Terminal condition\nZero silent omissions means:\n- every official ZIP identified;\n- every recursive manifestation terminally dispositioned;\n- all current-input bytes covered by source ledger;\n- all canonical objects source-bound;\n- all material conflicts resolved or explicitly blocking;\n- operational destinations/test evidence closed where applicable;\n- final packages and manual reconcile back to source manifestations.\n","createdAt",["D",1788774048394],"branchVersion","branchId","a63c93e7-3d77-4ab6-bcb1-01531faf38ba","id","3d35a1ad-763a-4f0a-b226-376a3cd33711","grimorio-evidence-memory",["D",1788770368230],"deletedAt","legacyEnvVars","privacy","public","httpPrivacy","description","Evidence ledger and semantic checkpoint engine","imageUrl","6cc24525-830b-4db3-9c9c-fd19fb50e91f","pinnedFileId","d572484c-aa99-11f1-bc26-1607ee4eb77e","pinPreview",true,"pinMigrated","screenshotVersion","likeCount",0,"orgId","01a0773d-9d6b-7580-a53a-683363c351c8","tursoName","3d35a1ad763a4f0ab226376a3cd33711","tursoHostname","3d35a1ad763a4f0ab226376a3cd33711-val-town.aws-us-east-2.turso.io","tursoMeasuredAt",["D",1789551045218],"pinned","userPerm","r","valPerm","blobPerm","rw","emailPerm","sqlitePerm","projectPerm","telemetryPerm","user",{"_98":97,"_106":102},"org",{"_98":97,"_99":100,"_101":102,"_18":53,"_103":104,"_105":24},"branch",{"_13":87,"_31":88,"_41":40,"_89":42,"_90":-5,"_91":-5,"_92":-5,"_93":-5,"_94":-5,"_95":43,"_96":97},"isOwned",149,"main","projectId","forkedBranchId","forkedBranchVersion","forkedProjectPrivacy","forkedProjectId","forkedProjectOrgId","projectName","userHandle","cleitoncosta","handle","displayName","Cleiton Costa","avatar","https://lh3.googleusercontent.com/a/ACg8ocIgWisWQiXts_2fy5iRiRgl8A6c7Ef94eqE3ur-4mlyUrpCBmQ=s96-c","tier","free","canRestrictVals","profileImageUrl","branchName","","adoptOrgId",{"_41":42,"_31":43,"_36":336,"_45":-5,"_46":24,"_47":48,"_49":48,"_50":51,"_52":-5,"_18":53,"_54":55,"_56":57,"_58":-5,"_59":-5,"_60":61,"_62":63,"_64":65,"_66":67,"_68":337,"_70":-5,"_71":72,"_73":72,"_74":75,"_76":75,"_77":75,"_78":72,"_79":72,"_80":338,"_82":339,"_84":340,"_86":24},"tree",[188,194,197,205,206,207],"files",[188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207],"counts",{"_178":61,"_179":149,"_180":61,"_181":182,"_183":184},"contents",[129,130,131,132,133,134],"upstreamProject","upstreamBranch","emptyEnvVars","flags",[],"hasHttpFiles","grants","orgs","customDomains",{},{"_41":172,"_31":173,"_29":174,"_137":-5,"_39":40,"_38":175,"_139":57,"_36":176,"_13":61,"_33":165,"_34":-5,"_23":24,"_144":-5,"_26":177},{"_41":167,"_31":168,"_29":169,"_137":-5,"_39":40,"_38":141,"_139":57,"_36":170,"_13":61,"_33":165,"_34":-5,"_23":24,"_144":-5,"_26":171},{"_41":160,"_31":161,"_29":162,"_137":-5,"_39":40,"_38":163,"_139":57,"_36":164,"_13":61,"_33":165,"_34":-5,"_23":24,"_144":-5,"_26":166},{"_41":153,"_31":154,"_29":155,"_137":-5,"_39":40,"_38":156,"_139":57,"_36":157,"_13":149,"_33":20,"_34":158,"_23":24,"_144":-5,"_26":159},{"_41":146,"_31":147,"_29":148,"_137":-5,"_39":40,"_38":149,"_139":57,"_36":150,"_13":61,"_33":20,"_34":151,"_23":24,"_144":-5,"_26":152},{"_41":135,"_31":136,"_29":55,"_137":-5,"_39":40,"_38":138,"_139":24,"_36":140,"_13":141,"_33":142,"_34":143,"_23":24,"_144":-5,"_26":145},"4f5a88c4-34ca-456d-bbbe-17ea00abf43b","api.ts","parentId",19,"latest",["D",1788773685578],2,"http","import { searchEvidence, unresolved } from \"./lib/search.ts\";\nimport { reconcile } from \"./lib/reconcile.ts\";\nimport { sqlite } from \"https://esm.town/v/std/sqlite/main.ts\";\n\nexport default async function (req: Request): Promise {\n const url = new URL(req.url);\n if (req.method !== \"GET\") return Response.json({error:\"READ_ONLY\"},{status:405});\n if (url.pathname.endsWith(\"/health\")) {\n const gates = await sqlite.execute(`SELECT gate_id,status,numerator,denominator,blocker,updated_at FROM completion_gates ORDER BY gate_id`);\n const cp = await sqlite.execute(`SELECT id,phase,status,created_at FROM checkpoints ORDER BY created_at DESC LIMIT 10`);\n return Response.json({status:\"PASS\",mode:\"READ_ONLY_METADATA_ONLY\",raw_source_exposed:false,gates:gates.rows??[],recent_checkpoints:cp.rows??[]});\n }\n if (url.pathname.endsWith(\"/unresolved\")) return Response.json(await unresolved(Math.min(Number(url.searchParams.get(\"limit\")||50),200)));\n if (url.pathname.endsWith(\"/reconcile\")) return Response.json(await reconcile());\n if (url.pathname.endsWith(\"/platforms\")) {\n const r=await sqlite.execute(`SELECT platform_id,instruction_surfaces,skill_surface,agent_surface,workflow_surface,mcp_surface,currentness_state,source_urls_json,notes,updated_at FROM platform_currentness ORDER BY platform_id`);\n return Response.json({status:\"PASS\",boundary:\"DOC_CURRENTNESS_NE_BINARY_AUTH_AUTHZ_QUOTA_TASK_PROOF\",platforms:r.rows??[]});\n }\n if (url.pathname.endsWith(\"/baselines\")) {\n const r=await sqlite.execute(`SELECT * FROM version_baselines ORDER BY date,baseline_id`);\n return Response.json({status:\"PASS\",role:\"NON_REGRESSION_REFERENCE_NOT_CURRENT_INPUT_AUTHORITY\",invariants:[\"CANONICAL_OBJECT_COUNT_NE_SOURCE_COVERAGE\",\"SEMANTIC_COMPRESSION_NE_PERMITTED_INFORMATION_LOSS\"],baselines:r.rows??[]});\n }\n const q=(url.searchParams.get(\"q\")||\"\").trim();\n if (!q) return Response.json({service:\"grimorio-evidence-memory\",raw_source_exposed:false,routes:[\"?q=term\",\"/health\",\"/unresolved\",\"/reconcile\",\"/platforms\",\"/baselines\"]});\n return Response.json({query:q,hits:await searchEvidence(q,Math.min(Number(url.searchParams.get(\"limit\")||25),100))});\n}\n","label",[136],"75787a32-0286-4915-9b97-82540ff09949","README.md","a3682d3c-aa97-11f1-8a9d-1607ee4eb77e",1,["D",1788770368285],"# Grimório Evidence Memory\n\nPublic engine code only. Raw Grimório source content is never stored in public files.\n\nThe project-scoped database stores hashes, states, semantic IDs, provenance pointers, checkpoints, and validation results. No raw source text is exposed by this val.\n",[147],"e45cf102-39c2-44fe-bfc5-2a0920659edb","CURRENT_STATE.md","abedfb0a-aa9a-11f1-9e1f-1607ee4eb77e",21,["D",1788773807563],"# Grimório 11-ZIP — Current State\n\nCheckpoint: `CP-20260907-CONTINUITY-005`\n\n## Input authority\n- Outer transport container received: `07092026_CAMADAS_GRIMORIO_NOVAS (2).zip`.\n- Official input set: **11 shallow ZIPs inside the outer container**.\n- Received: **11/11** by user declaration/transport structure.\n- Physically enumerated in current runtime: **0/11**.\n- Current blocker: the ChatGPT sandbox/file index cannot mount or enumerate the uploaded outer archive in this session.\n- `INPUT_SET_COMPLETE = BLOCKED`.\n- Final Research/Agentic/PDF/MD generation remains prohibited until physical census, full recursive read and reconciliation are proven.\n\n## External continuity/search layer\nVal Town project: `cleitoncosta/grimorio-evidence-memory`.\n\nPrivacy boundary: the project is on a Free Val Town account and its code is public. **No raw Grimório source text is stored there.** It stores only second-order metadata: hashes when known, states, provenance pointers, semantic IDs, findings, validation/checkpoint data and currentness references.\n\nVerified runtime surfaces:\n- Search engine: PASS.\n- `/health`: PASS.\n- `/unresolved`: PASS.\n- `/reconcile`: PASS as a fail-closed auditor; current verdict FAIL because terminal gates are not satisfied.\n- `/platforms`: PASS.\n- `/baselines`: implemented as non-regression reference endpoint.\n- Raw source exposure: false by contract.\n\nCurrent reconciler counts at this checkpoint:\n- 10 material/open findings in terminal reconciliation scope.\n- 5 unresolved/reference artifacts awaiting physical current-input binding.\n- 55 manual coverage requirements not yet PASS.\n- 6 semantic objects not yet terminally reconciled.\n\n## Historical non-regression baselines\nStructured `version_baselines` now includes:\n- V3 2026-08-29: 3,184 physical manifestations; 2,823 unique CONTENT_READ + 361 duplicate-with-reference; 657 deep indexed/routable Skills; 16 portable-core Skills; 16 logical Agents; parse-aware ingestion.\n- GUH V6: 131 canonical Skills; 31 specialized Agents; 29 always-on Rules; 839 unique objects; 30/30 theme coverage; 182 Markdown crosswalk; 7 adapters.\n- GUH V8.1 Living Harness: 5,036 current-input manifestations accounted; 4,986 text fully read; 50 binary byte-read; 558 physical Skill manifestations crosswalked; 242 canonical Skills + 15 aliases; 71 Agents; 148 Rules; 644 compact KB objects.\n- V9.2.1: 258 canonical Skills; 15 aliases; 60 active Agents; 175 active Rules; 373 glossary/taxonomy entries; universal operational component routing and deep-knowledge reachability.\n- R5 terminal full-content reference: 11,013 Research subject files / 11,026 Development subject files.\n\nThese are **version/scope-specific non-regression references**, not substitutes for the 11 current ZIPs.\n\n## V9.2.1 semantic contract recovered\nThe recovered contract applies semantic rigor to Skills, Rules, Agents, Validators, Tests, Templates, Routers, Retrievers, Registries, Knowledge Graphs, scripts, Generators, Runtime Governors, Ledgers, Adapters, Bootstrap, Release Gates, Manifests, Schemas, Config, CI/CD, Migrations, Fixtures, Eval Datasets, Benchmarks, Prompts, System Instructions, Policies, Crosswalks, Receipts, Reports, Generated/Status Docs, Capability Catalogs, Aliases, Dependency Maps, Recovery, Installers, Packaging, Compatibility Matrices, Run State, State Machines, Concurrency Controls, Tool/Plugin Metadata, Source Provenance and Negative Knowledge.\n\nCore invariant:\n`REQUIRED_KNOWLEDGE ⊆ LOCAL_KNOWLEDGE ∪ REFERENCED_KNOWLEDGE ∪ ROUTABLE_KNOWLEDGE`\n\nSelf-application must close both directions:\n- source/concept -> operational component/consumer/enforcement/test/mutation/evidence;\n- active component -> source knowledge/contract/consumer/trigger/output/enforcement/detector/evidence.\n\n## Platform currentness — 2026-09-07\nCurrent documentation was recensused for:\n- Agent Skills open standard;\n- OpenAI Codex;\n- Anthropic Claude Code;\n- Gemini CLI;\n- GitHub Copilot CLI;\n- Cursor;\n- Google Antigravity 2.0;\n- Devin Desktop / Windsurf lineage.\n\nFour older/partial adapter records (`openai`, `anthropic-general`, `google-antigravity`, `windsurf`) were moved to `platform_currentness_history` with `superseded_by` instead of being silently deleted.\n\nImportant current facts:\n- Codex: AGENTS hierarchy, Skills, MCP and Subagents are distinct surfaces.\n- Claude Code: CLAUDE.md, Skills, Hooks, MCP and Subagents are distinct; hooks are lifecycle enforcement rather than prompt-only guidance.\n- Gemini CLI: GEMINI.md, Skills, Hooks, Policies and Subagents are distinct. Current official docs state workspace policy tier is disabled/non-functional; adapter must not emit it as effective enforcement.\n- Copilot CLI: multiple instruction sources are combined and no general precedence is documented across all; do not invent precedence.\n- Cursor: `.cursor/rules` requires `.mdc`; plain `.md` there is ignored. AGENTS.md and Skills/Subagents have separate semantics.\n- Devin/Windsurf lineage: Skills, Rules and Workflows are explicitly distinct.\n- Antigravity 2.0: Projects/worktrees, scoped permissions, Rules/Workflows, JSON Hooks, Skills, MCP, Subagents and Artifacts are distinct surfaces.\n\nBoundary:\n`DOC_CURRENTNESS != BINARY_INSTALLED != CONNECTED != AUTHENTICATED != AUTHORIZED != ENTITLED/QUOTA != VERIFIED_FOR_TASK`.\n\nFirecrawl in current session: connector surface exists but search returned HTTP 402 -> `QUOTA_BLOCKED`, not unavailable.\n\n## Manual architecture\n55 coverage requirements are registered. Canonical Markdown will be the single semantic source for PDF generation. Information architecture is book-like sequential learning plus explicit Explanation/Tutorial/How-to/Reference modes and Atlas/Glossary lookup.\n\n## Critical invariants\n`NO_TEXT_HIT != ABSENCE`\n`FOUND != READ != APPLIED != TESTED`\n`SKILL_EXISTS != SKILL_APPLIED`\n`AGENT_DEFINED != AGENT_EXECUTED`\n`AUTHENTICATED != AUTHORIZED`\n`TEST_PASS != CURRENT_EVIDENCE_PASS`\n`ASSURANCE_COMPLETE != CORPUS_COMPLETE`\n`CANONICAL_OBJECT_COUNT != SOURCE_COVERAGE`\n`SEMANTIC_COMPRESSION != PERMITTED_INFORMATION_LOSS`\n`NO_LOSSY_OPERATIONAL_COMPRESSION`\n`KNOWLEDGE_REACHABILITY`\n`PLATFORM_FILE_FORMAT != SEMANTIC_ROLE`\n\n## Current next atomic action\nRetry physical archive mount/indexing periodically without replacing it with historical inference. Once accessible:\n\n`outer ZIP -> exactly 11 shallow ZIPs -> SHA-256/size/CRC/archive safety -> recursive nested-archive census -> manifestation-level ledger -> actual-format + encoding-aware read -> exact/semantic duplicate graph -> semantic reconciliation -> current adapter generation -> implementation/test/mutation/currentness lifecycle`.\n\nEvery physical source manifestation must end with an explicit disposition. A smaller canonical object count is acceptable only when source coverage, provenance and knowledge reachability remain complete.\n",[154],"f83fe8d5-f32c-4a5c-a50a-75747e69b154","scripts","d4184124-aa97-11f1-b2c7-1607ee4eb77e",3,["D",1788770450014],"directory",[161],"456c8ca9-da5c-4255-9a54-f249553707e7","lib","cff173e0-aa97-11f1-b2c7-1607ee4eb77e",["D",1788770443012],[168],"4a32415c-ea58-4d90-b7d9-8d0c77b2da49","docs","d76e827a-aa97-11f1-aeb6-1607ee4eb77e",4,["D",1788770455617],[173],"pulls","branches","forks","envVars",{"_185":61,"_187":61},"connections",{"_185":61,"_186":61},"count","unfilledCount","emptyCount",{"_41":172,"_31":173,"_29":174,"_137":-5,"_39":40,"_38":175,"_139":57,"_36":333,"_13":61,"_33":165,"_34":-5,"_23":24,"_144":-5,"_209":334,"_26":335},{"_41":325,"_31":326,"_29":327,"_137":174,"_39":40,"_38":328,"_139":57,"_36":329,"_13":61,"_33":20,"_34":330,"_23":24,"_144":-5,"_209":331,"_26":332},{"_41":321,"_31":32,"_29":30,"_137":174,"_39":40,"_38":14,"_139":57,"_36":322,"_13":61,"_33":20,"_34":35,"_23":24,"_144":-5,"_209":323,"_26":324},{"_41":313,"_31":314,"_29":315,"_137":174,"_39":40,"_38":316,"_139":57,"_36":317,"_13":61,"_33":20,"_34":318,"_23":24,"_144":-5,"_209":319,"_26":320},{"_41":305,"_31":306,"_29":307,"_137":174,"_39":40,"_38":308,"_139":57,"_36":309,"_13":61,"_33":20,"_34":310,"_23":24,"_144":-5,"_209":311,"_26":312},{"_41":297,"_31":298,"_29":299,"_137":174,"_39":40,"_38":300,"_139":57,"_36":301,"_13":61,"_33":20,"_34":302,"_23":24,"_144":-5,"_209":303,"_26":304},{"_41":167,"_31":168,"_29":169,"_137":-5,"_39":40,"_38":141,"_139":57,"_36":294,"_13":61,"_33":165,"_34":-5,"_23":24,"_144":-5,"_209":295,"_26":296},{"_41":286,"_31":287,"_29":288,"_137":169,"_39":40,"_38":289,"_139":24,"_36":290,"_13":149,"_33":223,"_34":291,"_23":24,"_144":-5,"_209":292,"_26":293},{"_41":278,"_31":279,"_29":280,"_137":169,"_39":40,"_38":281,"_139":24,"_36":282,"_13":61,"_33":223,"_34":283,"_23":24,"_144":-5,"_209":284,"_26":285},{"_41":160,"_31":161,"_29":162,"_137":-5,"_39":40,"_38":163,"_139":57,"_36":275,"_13":61,"_33":165,"_34":-5,"_23":24,"_144":-5,"_209":276,"_26":277},{"_41":267,"_31":268,"_29":269,"_137":162,"_39":40,"_38":270,"_139":24,"_36":271,"_13":61,"_33":223,"_34":272,"_23":24,"_144":-5,"_209":273,"_26":274},{"_41":259,"_31":260,"_29":261,"_137":162,"_39":40,"_38":262,"_139":57,"_36":263,"_13":61,"_33":223,"_34":264,"_23":24,"_144":-5,"_209":265,"_26":266},{"_41":251,"_31":252,"_29":253,"_137":162,"_39":40,"_38":254,"_139":57,"_36":255,"_13":149,"_33":223,"_34":256,"_23":24,"_144":-5,"_209":257,"_26":258},{"_41":243,"_31":244,"_29":245,"_137":162,"_39":40,"_38":246,"_139":57,"_36":247,"_13":61,"_33":223,"_34":248,"_23":24,"_144":-5,"_209":249,"_26":250},{"_41":235,"_31":236,"_29":237,"_137":162,"_39":40,"_38":238,"_139":57,"_36":239,"_13":61,"_33":223,"_34":240,"_23":24,"_144":-5,"_209":241,"_26":242},{"_41":227,"_31":228,"_29":229,"_137":162,"_39":40,"_38":230,"_139":57,"_36":231,"_13":61,"_33":223,"_34":232,"_23":24,"_144":-5,"_209":233,"_26":234},{"_41":218,"_31":219,"_29":220,"_137":162,"_39":40,"_38":221,"_139":57,"_36":222,"_13":61,"_33":223,"_34":224,"_23":24,"_144":-5,"_209":225,"_26":226},{"_41":153,"_31":154,"_29":155,"_137":-5,"_39":40,"_38":156,"_139":57,"_36":215,"_13":149,"_33":20,"_34":158,"_23":24,"_144":-5,"_209":216,"_26":217},{"_41":146,"_31":147,"_29":148,"_137":-5,"_39":40,"_38":149,"_139":57,"_36":212,"_13":61,"_33":20,"_34":151,"_23":24,"_144":-5,"_209":213,"_26":214},{"_41":135,"_31":136,"_29":55,"_137":-5,"_39":40,"_38":138,"_139":24,"_36":208,"_13":141,"_33":142,"_34":143,"_23":24,"_144":-5,"_209":210,"_26":211},["D",1788773685578],"children",[],[136],["D",1788770368285],[],[147],["D",1788773807563],[],[154],"bb2f7a13-b000-49ff-8fc1-5305e6f751c6","seed_quality_requirements.ts","080b8934-aaa0-11f1-aea4-1607ee4eb77e",23,["D",1788773973125],"script","import { sqlite } from \"https://esm.town/v/std/sqlite/main.ts\";\n\nconst falsifier = [\n\"Every relevant learning has an operational destination when necessary\",\n\"Every active component has justifiable source/contract/consumer\",\n\"No active file is validated only by presence/name\",\n\"Manual/current authority is current and coherent with code/registries\",\n\"No generator can silently recreate stale documentation\",\n\"No active template depends on uncontracted placeholders\",\n\"No canonical Skill is nominal boilerplate\",\n\"No independent Agent exists without context/authority justification\",\n\"No machine-checkable Rule is claimed enforced without a detector\",\n\"Every critical validator detects a faithful mutation/reproducer\",\n\"No critical test uses the implementation under test as its own oracle\",\n\"Router generalizes beyond keywords and development examples\",\n\"Retrieval uses every layer claimed by documentation or the claim is reduced\",\n\"Operational graphs have a real consumer and influence mutation\",\n\"Portable compiles pre-project intelligence from the full GUH rather than a narrow subset\",\n\"Chatbot discovers runtime capabilities and installable providers when materially needed\",\n\"Missing tool/plugin capability produces explicit gap/fallback rather than silent degradation\",\n\"Artifact-first research reaches primary artifacts when pertinent\",\n\"Compute providers are used as compute/correlation rather than source authority\",\n\"Knowledge Graph is navigation rather than truth authority\",\n\"Low resource budget does not silently reduce quality\",\n\"Partial staging cannot receive candidate/final receipt\",\n\"All runtime agents are accounted for and terminated/cancelled\",\n\"No unexpected orphan background resource remains after closure\",\n\"Parallelism is selected for benefit rather than mere availability\",\n\"Progressive disclosure is applied to the Grimório itself\",\n\"Irrelevant tools are not permanently loaded\",\n\"Executables do not hide placeholder/pass/simulation claimed as runtime\",\n\"Side effects/writers have ownership, idempotency and cleanup\",\n\"The Grimório applies its own rules to itself\"\n];\n\nconst resourceMutations = [\n\"quota ends with package 70 percent generated\",\n\"credit ends mid Skill or Markdown generation\",\n\"credit ends before validation reserve\",\n\"parent terminates while child continues\",\n\"child ignores cancellation\",\n\"heartbeat remains alive without progress\",\n\"recursive fan-out exceeds governance\",\n\"duplicate agent runs same task without information-gain justification\",\n\"worker calls tool after root run COMPLETE\",\n\"partial package attempts to receive final receipt\",\n\"writer crashes mid JSON/CSV and leaves partial authority\",\n\"identical retry occurs after deterministic quota/entitlement failure\"\n];\n\nawait sqlite.execute(`DELETE FROM quality_requirements WHERE family IN ('FINAL_FALSIFIER','RESOURCE_MUTATION')`);\nlet i=0;\nfor (const r of falsifier) {\n i++;\n await sqlite.execute({sql:`INSERT INTO quality_requirements(id,family,requirement,status,updated_at) VALUES (?,?,?,?,datetime('now'))`,args:[`FF-${String(i).padStart(3,'0')}`,'FINAL_FALSIFIER',r,'PENDING']});\n}\ni=0;\nfor (const r of resourceMutations) {\n i++;\n await sqlite.execute({sql:`INSERT INTO quality_requirements(id,family,requirement,status,updated_at) VALUES (?,?,?,?,datetime('now'))`,args:[`RM-${String(i).padStart(3,'0')}`,'RESOURCE_MUTATION',r,'PENDING']});\n}\nconst q=await sqlite.execute(`SELECT family,COUNT(*) AS n,SUM(CASE WHEN status='PASS' THEN 1 ELSE 0 END) AS pass FROM quality_requirements GROUP BY family ORDER BY family`);\nconsole.log(JSON.stringify({status:'PASS',families:q.rows??[]},null,2));\n",[],[161,219],"9ebbc449-16ad-4054-902d-7709b5e6611a","seed_manual_requirements.ts","6274a2ee-aa9a-11f1-a6b5-1607ee4eb77e",11,["D",1788771547843],"import { sqlite } from \"https://esm.town/v/std/sqlite/main.ts\";\nconst req = [\n[\"M001\",\"Sistemas de IA\",\"Computação, algoritmos e dados\"],[\"M002\",\"Sistemas de IA\",\"Estatística, probabilidade e avaliação\"],[\"M003\",\"Sistemas de IA\",\"Machine learning e deep learning\"],[\"M004\",\"Sistemas de IA\",\"Redes neurais, embeddings, transformers e attention\"],[\"M005\",\"Sistemas de IA\",\"Tokenização, context window e inferência\"],[\"M006\",\"Sistemas de IA\",\"Training, fine-tuning e adaptation\"],[\"M007\",\"Sistemas de IA\",\"Prompting e context engineering\"],[\"M008\",\"Sistemas de IA\",\"RAG, retrieval, vector databases e knowledge graphs\"],[\"M009\",\"Sistemas de IA\",\"Memory, state e context durability\"],[\"M010\",\"Sistemas de IA\",\"Tools, function calling e MCP\"],[\"M011\",\"Sistemas de IA\",\"Agents, subagents e multi-agent systems\"],[\"M012\",\"Sistemas de IA\",\"Orchestration, routers, planners e handoffs\"],[\"M013\",\"Sistemas de IA\",\"Validators, evaluators, graders e guardrails\"],[\"M014\",\"Sistemas de IA\",\"Hallucination, uncertainty, provenance e currentness\"],[\"M015\",\"Sistemas de IA\",\"Observability, traces, evals e regression\"],[\"M016\",\"Sistemas de IA\",\"Safety, alignment, security e prompt injection\"],[\"M017\",\"Sistemas de IA\",\"Inference infrastructure, GPUs, local e cloud\"],[\"M018\",\"Desenvolver com IA\",\"Ideação, problema e pesquisa\"],[\"M019\",\"Desenvolver com IA\",\"Requirements, PRD e build spec\"],[\"M020\",\"Desenvolver com IA\",\"Context engineering e token economy\"],[\"M021\",\"Desenvolver com IA\",\"Skills, Agents, Rules e instruction surfaces\"],[\"M022\",\"Desenvolver com IA\",\"Tool routing, capability discovery e currentness\"],[\"M023\",\"Desenvolver com IA\",\"Planning e task decomposition\"],[\"M024\",\"Desenvolver com IA\",\"Coding agents e implementation workflow\"],[\"M025\",\"Desenvolver com IA\",\"Testing, debugging e code review\"],[\"M026\",\"Desenvolver com IA\",\"Security e side-effect semantics\"],[\"M027\",\"Desenvolver com IA\",\"CI/CD, deployment e runtime verification\"],[\"M028\",\"Desenvolver com IA\",\"Checkpoints, recovery e durable execution\"],[\"M029\",\"Desenvolver com IA\",\"Failure mining e success mining\"],[\"M030\",\"Desenvolver com IA\",\"Mutation testing, falsification e clean-room\"],[\"M031\",\"Engenharia de Software\",\"Problema, domínio e requisitos\"],[\"M032\",\"Engenharia de Software\",\"Arquitetura e trade-offs\"],[\"M033\",\"Engenharia de Software\",\"Frontend e experiência do usuário\"],[\"M034\",\"Engenharia de Software\",\"Backend e serviços\"],[\"M035\",\"Engenharia de Software\",\"Bancos de dados, SQL e NoSQL\"],[\"M036\",\"Engenharia de Software\",\"APIs, REST, GraphQL e RPC\"],[\"M037\",\"Engenharia de Software\",\"Autenticação, autorização e segurança\"],[\"M038\",\"Engenharia de Software\",\"Redes, protocolos, storage, cache, filas e eventos\"],[\"M039\",\"Engenharia de Software\",\"Monolith, modular monolith e microservices\"],[\"M040\",\"Engenharia de Software\",\"Clean Architecture, DDD, SOLID e patterns\"],[\"M041\",\"Engenharia de Software\",\"Testing pyramid, unit, integration, E2E e contract tests\"],[\"M042\",\"Engenharia de Software\",\"Performance, scalability e observability\"],[\"M043\",\"Engenharia de Software\",\"Git, branches, commits e pull requests\"],[\"M044\",\"Engenharia de Software\",\"CI, CD, containers e cloud\"],[\"M045\",\"Engenharia de Software\",\"Environments, secrets, config e infrastructure as code\"],[\"M046\",\"Engenharia de Software\",\"Deployment, rollback, maintenance e technical debt\"],[\"M047\",\"Engenharia de Software\",\"Árvores de projeto e organização de repositório\"],[\"M048\",\"Engenharia de Software\",\"Tipos de arquivos e formatos\"],[\"M049\",\"Referência transversal\",\"Por que A e não B: decisões e trade-offs\"],[\"M050\",\"Referência transversal\",\"Glossário mestre, aliases e terminologia profissional\"],[\"M051\",\"Referência transversal\",\"Laboratórios, exemplos, erros e contrafactuais\"],[\"M052\",\"Referência transversal\",\"Plataformas de agentes e adapters atuais\"],[\"M053\",\"Referência transversal\",\"Provenance, source ledger e evidência\"],[\"M054\",\"Referência transversal\",\"Portabilidade Windows/Linux/macOS\"],[\"M055\",\"Referência transversal\",\"Acessibilidade, clareza e linguagem inclusiva\"]\n];\nconst sections=JSON.stringify([\"Introdução\",\"Motivação\",\"Conceitos\",\"Mecanismo\",\"Exemplos\",\"Trade-offs\",\"Falhas comuns\",\"Boas práticas\",\"Resumo\",\"Termos deste capítulo\",\"Conexões\"]);\nfor (const [id,family,topic] of req) await sqlite.execute({sql:`INSERT OR REPLACE INTO manual_coverage_requirements(id,family,topic,required_sections,source_basis,status,destination,coverage_evidence,updated_at) VALUES(?,?,?,?,?,?,?,?,datetime('now'))`,args:[id,family,topic,sections,\"USER_MASTER_PROMPT+CURRENT_STYLE_RESEARCH\",\"REQUIREMENT_ONLY\",null,null]});\nconsole.log(JSON.stringify({status:\"PASS\",requirements:req.length},null,2));\n",[],[161,228],"01c14c81-866a-4b93-b454-2ea55ed54a6c","seed_current_baselines.ts","35a404b2-aa9a-11f1-9220-1607ee4eb77e",10,["D",1788771472653],"import { sqlite } from \"https://esm.town/v/std/sqlite/main.ts\";\n\nconst topics = [\n[\"A001\",\"State durability\",\"YES\",\"HIGH\",\"HIGH\",\"CRITICAL\",\"MERGE\"],[\"A002\",\"Evidence typing\",\"YES\",\"HIGH\",\"HIGH\",\"CRITICAL\",\"MERGE\"],[\"A003\",\"Currentness DAG\",\"YES\",\"HIGH\",\"HIGH\",\"CRITICAL\",\"MERGE\"],[\"A004\",\"Mutation discriminance\",\"YES\",\"MEDIUM\",\"HIGH\",\"CRITICAL\",\"MERGE\"],[\"A005\",\"Independent validator\",\"YES\",\"HIGH\",\"HIGH\",\"CRITICAL\",\"MERGE\"],[\"A006\",\"Versioned memory\",\"YES\",\"HIGH\",\"HIGH\",\"CRITICAL\",\"MERGE\"],[\"A007\",\"Control surface/payload/evidence split\",\"YES\",\"MEDIUM\",\"HIGH\",\"CRITICAL\",\"MERGE\"],[\"A008\",\"Contradiction agent\",\"NO\",\"HIGH\",\"LOW\",\"HIGH\",\"RESEARCH_ONLY\"],[\"A009\",\"Citation audit\",\"NO\",\"HIGH\",\"LOW\",\"HIGH\",\"RESEARCH_ONLY\"],[\"A010\",\"Source currentness monitor\",\"NO\",\"HIGH\",\"MEDIUM\",\"HIGH\",\"RESEARCH_ONLY\"],[\"A011\",\"Single-agent-first\",\"NO\",\"MEDIUM\",\"HIGH\",\"HIGH\",\"AGENTIC_ONLY\"],[\"A012\",\"Semantic transactions\",\"NO\",\"LOW\",\"HIGH\",\"HIGH\",\"AGENTIC_ONLY\"],[\"A013\",\"Process-tree safety\",\"NO\",\"LOW\",\"HIGH\",\"HIGH\",\"AGENTIC_ONLY\"],[\"A014\",\"Bounded output\",\"NO\",\"LOW\",\"HIGH\",\"HIGH\",\"AGENTIC_ONLY\"],[\"A015\",\"Durable execution\",\"NO\",\"LOW\",\"HIGH\",\"HIGH\",\"AGENTIC_ONLY\"],[\"A016\",\"MCP/A2A boundary\",\"NO\",\"MEDIUM\",\"HIGH\",\"HIGH\",\"AGENTIC_ONLY\"],[\"A017\",\"OpenTelemetry observability\",\"YES\",\"MEDIUM\",\"HIGH\",\"HIGH\",\"MERGE\"],[\"A018\",\"Trace→Eval→Regression\",\"YES\",\"HIGH\",\"HIGH\",\"CRITICAL\",\"MERGE\"]\n];\nfor (const r of topics) await sqlite.execute({sql:`INSERT OR REPLACE INTO application_topics(id,topic,shared,research_weight,agentic_weight,priority,action,source_state,provenance_json) VALUES(?,?,?,?,?,?,?,?,?)`,args:[...r,\"LIBRARY_DERIVED_PARTIAL\",JSON.stringify({source:\"APPLICATION_MATRIX.csv\",file_ref:\"turn81file1\"})]});\nconst failures = [\n[\"F001\",\"CONVERSATION_AS_STATE\",\"Conversa usada como estado autoritativo\",\"checkpoint externo + JSON + hashes\",\"interromper e reconstruir\"],[\"F002\",\"HISTORICAL_PASS_PROMOTED\",\"PASS histórico tratado como atual\",\"currentness bindings + evidence typing\",\"mutar runner/control surface\"],[\"F003\",\"WRONG_DETECTOR_MUTATION\",\"Mutation morre pelo detector errado\",\"expected_detector obrigatório\",\"dependency sem rebind\"],[\"F004\",\"UNEXPECTED_EXCEPTION_AS_KILL\",\"Exceção do falsifier contabilizada como kill\",\"unexpected exception = campaign failure\",\"forçar exception\"],[\"F005\",\"PARTIAL_ARTIFACT_LEAK\",\"Timeout deixa ZIP parcial\",\"quarantine + manifest\",\"interromper build\"],[\"F006\",\"PROCESS_CHILD_ESCAPE\",\"Filho sobrevive ao timeout\",\"process group/session kill\",\"spawn child atrasado\"],[\"F007\",\"UNBOUNDED_OUTPUT\",\"Runner gera saída ilimitada\",\"bounded output\",\"stdout flood\"],[\"F008\",\"SOURCE_DRIFT_ON_RESUME\",\"Resume após source mudar\",\"source digest checkpoint\",\"editar source entre runs\"],[\"F009\",\"SYMLINK_ARCHIVE_ESCAPE\",\"Path/symlink escapa intake\",\"safe archive intake\",\"traversal/symlink ZIP\"],[\"F010\",\"DAG_EMPTY_FALSE_PASS\",\"DAG vazio retorna PASS\",\"minimum-node/non-empty gate\",\"DAG vazio\"],[\"F011\",\"CROSS_PACKAGE_EMPTY_SET\",\"shared set vazio gera PASS\",\"non-empty semantic set\",\"shared=[]\"],[\"F012\",\"RECEIPT_REPLAY\",\"receipt stale/externo reutilizado\",\"full fingerprint binding\",\"reusar após mudança\"],[\"F013\",\"SELF_REFERENTIAL_DIGEST\",\"evidence contamina digest do target\",\"namespaces separados\",\"gerar report dentro target\"],[\"F014\",\"NAME_HEURISTIC_EXCLUSION\",\"arquivo material excluído por nome\",\"namespace classification\",\"safety_report.json\"],[\"F015\",\"MEMORY_PERSISTENCE_ASSUMED\",\"compiled memory tratada como persisted\",\"compiled/persisted/recall separated\",\"novo contexto\"],[\"F016\",\"NO_HIT_AS_ABSENCE\",\"busca sem hit vira inexistência\",\"negative-space retrieval\",\"anexos/versions/refs\"],[\"F017\",\"OVER_AGENTIFICATION\",\"subagentes sem contrato distinto\",\"single-agent-first\",\"ablation agent count\"],[\"F018\",\"NON_IDEMPOTENT_RETRY\",\"retry duplica efeito externo\",\"idempotency key + effect ledger\",\"retry after partial commit\"],[\"F019\",\"TOOL_HTTP_SUCCESS_FALSE_GREEN\",\"HTTP 200 tratado como sucesso sem state check\",\"postcondition validation\",\"mock 200 wrong state\"],[\"F020\",\"LONG_CONTEXT_ACCUMULATION\",\"contexto cresce e perde sinal\",\"typed lifecycle + external evidence memory\",\"long-horizon replay\"],[\"F021\",\"MEMORY_POISONING\",\"memória persistente aceita conteúdo adversarial\",\"provenance + trust + rollback\",\"inject stale/malicious memory\"],[\"F022\",\"GRADER_HACKING\",\"sistema otimiza para validador fraco\",\"independent/human/deterministic cross-check\",\"adversarial eval\"],[\"F023\",\"SIDE_EFFECT_CLASS_UNKNOWN\",\"ação mutável executada sem classificação\",\"pre-execution effect declaration\",\"mutating tool mislabeled\"],[\"F024\",\"CAPABILITY_CATALOG_ASSUMED\",\"plugin listado tratado como executável\",\"runtime census\",\"quota/auth failure\"]\n];\nfor (const r of failures) await sqlite.execute({sql:`INSERT OR REPLACE INTO failure_class_catalog(id,class_name,observation,control,regression,source_state,provenance_json) VALUES(?,?,?,?,?,?,?)`,args:[...r,\"LIBRARY_DERIVED_PARTIAL\",JSON.stringify({source:\"FAILURE_CLASSES.csv\",file_ref:\"turn81file0\"})]});\nconsole.log(JSON.stringify({status:\"PASS\",application_topics:topics.length,failure_classes:failures.length},null,2));\n",[],[161,236],"f561e46b-e265-4174-b3cd-baf67c0dafca","search_smoke.ts","63a17efa-aa98-11f1-8444-1607ee4eb77e",8,["D",1788770690811],"import { searchEvidence, unresolved } from \"../lib/search.ts\";\nconst a = await searchEvidence(\"reachability\", 10);\nconst b = await searchEvidence(\"capability\", 10);\nconst u = await unresolved(10);\nconsole.log(JSON.stringify({status:(a.length>0 && b.length>0)?\"PASS\":\"FAIL\", reachability:a, capability:b, unresolved:u},null,2));\n",[],[161,244],"5446b63e-f180-4456-ae1c-db885d05ca67","reconciliation_gate.ts","b5a298e0-aa9a-11f1-8b92-1607ee4eb77e",17,["D",1788773598555],"import { reconcile } from \"../lib/reconcile.ts\";\n\nconst out = await reconcile();\nconsole.log(JSON.stringify(out,null,2));\nif (out.status !== \"PASS\") Deno.exit(2);\n",[],[161,252],"290a7fec-201c-4f4b-ad96-1a45bf0b3f56","non_regression_gate.ts","b6904432-aa9f-11f1-8aa4-1607ee4eb77e",22,["D",1788773836406],"import { sqlite } from \"https://esm.town/v/std/sqlite/main.ts\";\n\nconst current = await sqlite.execute(`SELECT * FROM inventory_snapshots WHERE scope IN ('CURRENT_INPUT','FINAL_RESEARCH','FINAL_DEVELOPMENT') ORDER BY created_at DESC`);\nif (!(current.rows?.length)) {\n console.log(JSON.stringify({\n schema:\"grimorio.non-regression-gate/1.0\",\n status:\"BLOCKED\",\n reasons:[\"NO_CURRENT_PHYSICAL_INVENTORY_SNAPSHOT\"],\n invariant:\"CARDINALITY_IS_A_SIGNAL_NOT_A_COMPLETENESS_ORACLE\"\n },null,2));\n Deno.exit(2);\n}\n\nconst baselines = await sqlite.execute(`SELECT * FROM version_baselines ORDER BY date,baseline_id`);\nconst results=[];\nfor (const s of current.rows as any[]) {\n const reasons:string[]=[];\n const warnings:string[]=[];\n const manifestations=Number(s.manifestations ?? 0);\n const dispositioned=Number(s.dispositioned_manifestations ?? 0);\n if (!manifestations) reasons.push(\"ZERO_OR_UNKNOWN_MANIFESTATIONS\");\n if (manifestations && dispositioned !== manifestations) reasons.push(`SOURCE_DISPOSITION_GAP:${dispositioned}/${manifestations}`);\n if (String(s.provenance_coverage_state)!==\"PASS\") reasons.push(`PROVENANCE_COVERAGE:${s.provenance_coverage_state}`);\n if (String(s.semantic_reachability_state)!==\"PASS\") reasons.push(`SEMANTIC_REACHABILITY:${s.semantic_reachability_state}`);\n const maxSkill=657;\n const skillCount=Number(s.canonical_skills ?? 0);\n if (skillCount && skillCount < maxSkill) {\n warnings.push(`CANONICAL_SKILL_COUNT_BELOW_HISTORICAL_DEEP_ROUTABLE_FLOOR:${skillCount}<${maxSkill}`);\n if (String(s.compression_justification_state)!==\"PASS\") reasons.push(\"SKILL_CARDINALITY_SHRINK_WITHOUT_PROVEN_COMPRESSION_JUSTIFICATION\");\n }\n const rfiles=Number(s.research_files ?? 0);\n const dfiles=Number(s.development_files ?? 0);\n if (s.scope===\"FINAL_RESEARCH\" && rfiles && rfiles<11013) warnings.push(`RESEARCH_FILE_COUNT_BELOW_R5_REFERENCE:${rfiles}<11013`);\n if (s.scope===\"FINAL_DEVELOPMENT\" && dfiles && dfiles<11026) warnings.push(`DEVELOPMENT_FILE_COUNT_BELOW_R5_REFERENCE:${dfiles}<11026`);\n if ((warnings.length>0) && String(s.compression_justification_state)!==\"PASS\") reasons.push(\"VERSIONED_CARDINALITY_REGRESSION_NOT_JUSTIFIED\");\n results.push({snapshot_id:s.snapshot_id,scope:s.scope,status:reasons.length?\"FAIL\":\"PASS\",reasons,warnings});\n}\nconst out={schema:\"grimorio.non-regression-gate/1.0\",status:results.every((r:any)=>r.status===\"PASS\")?\"PASS\":\"FAIL\",results,baseline_count:baselines.rows?.length??0,invariants:[\"CANONICAL_OBJECT_COUNT_NE_SOURCE_COVERAGE\",\"SEMANTIC_COMPRESSION_NE_PERMITTED_INFORMATION_LOSS\",\"CARDINALITY_IS_A_SIGNAL_NOT_A_COMPLETENESS_ORACLE\"]};\nconsole.log(JSON.stringify(out,null,2));\nif (out.status!==\"PASS\") Deno.exit(2);\n",[],[161,260],"4abc8f0f-a767-4fc6-abd1-cd4a0b3f317f","health.ts","eb088060-aa97-11f1-8436-1607ee4eb77e",7,["D",1788770488489],"import { sqlite } from \"https://esm.town/v/std/sqlite/main.ts\";\n\nconst tables = [\"source_packages\",\"source_artifacts\",\"semantic_objects\",\"semantic_relations\",\"findings\",\"validations\",\"checkpoints\",\"implementation_ledger\",\"completion_gates\"];\nconst counts: Record = {};\nfor (const table of tables) {\n const r = await sqlite.execute(`SELECT COUNT(*) AS n FROM ${table}`);\n counts[table] = Number(r.rows?.[0]?.n ?? 0);\n}\nconst gates = await sqlite.execute(`SELECT gate_id,status,numerator,denominator,blocker FROM completion_gates ORDER BY gate_id`);\nconsole.log(JSON.stringify({status:\"PASS\",counts,gates:gates.rows ?? []},null,2));\n",[],[161,268],["D",1788770450014],[198,199,200,201,202,203,204],[161],"16e8e144-d335-416a-b096-f76af70f172a","search.ts","e6a4c718-aa97-11f1-bb39-1607ee4eb77e",6,["D",1788770481128],"import { sqlite } from \"https://esm.town/v/std/sqlite/main.ts\";\n\nexport type SearchHit = {\n kind: \"artifact\" | \"semantic\" | \"finding\" | \"checkpoint\";\n id: string;\n label: string;\n state: string;\n score: number;\n};\n\nfunction escLike(q: string) {\n return `%${q.replaceAll(\"%\", \"\\\\%\").replaceAll(\"_\", \"\\\\_\")}%`;\n}\n\nexport async function searchEvidence(query: string, limit = 25): Promise {\n const q = escLike(query.trim().toLowerCase());\n const hits: SearchHit[] = [];\n const semantic = await sqlite.execute({\n sql: `SELECT id, canonical_name, status FROM semantic_objects WHERE lower(canonical_name) LIKE ? ESCAPE '\\\\' OR lower(COALESCE(definition,'')) LIKE ? ESCAPE '\\\\' LIMIT ?`,\n args: [q, q, limit],\n });\n for (const row of semantic.rows ?? []) hits.push({kind:\"semantic\", id:String(row.id), label:String(row.canonical_name), state:String(row.status), score:100});\n\n const findings = await sqlite.execute({\n sql: `SELECT id, finding_class, status FROM findings WHERE lower(finding_class) LIKE ? ESCAPE '\\\\' OR lower(observation) LIKE ? ESCAPE '\\\\' LIMIT ?`,\n args: [q, q, limit],\n });\n for (const row of findings.rows ?? []) hits.push({kind:\"finding\", id:String(row.id), label:String(row.finding_class), state:String(row.status), score:90});\n\n const artifacts = await sqlite.execute({\n sql: `SELECT id, original_path, read_state FROM source_artifacts WHERE lower(original_path) LIKE ? ESCAPE '\\\\' LIMIT ?`,\n args: [q, limit],\n });\n for (const row of artifacts.rows ?? []) hits.push({kind:\"artifact\", id:String(row.id), label:String(row.original_path), state:String(row.read_state), score:80});\n\n const checkpoints = await sqlite.execute({\n sql: `SELECT id, phase, status FROM checkpoints WHERE lower(id) LIKE ? ESCAPE '\\\\' OR lower(phase) LIKE ? ESCAPE '\\\\' LIMIT ?`,\n args: [q, q, limit],\n });\n for (const row of checkpoints.rows ?? []) hits.push({kind:\"checkpoint\", id:String(row.id), label:String(row.phase), state:String(row.status), score:70});\n\n return hits.sort((a,b) => b.score-a.score || a.label.localeCompare(b.label)).slice(0, limit);\n}\n\nexport async function unresolved(limit = 100) {\n const artifacts = await sqlite.execute({\n sql: `SELECT id, package_id, original_path, read_state, semantic_state, disposition_state FROM source_artifacts WHERE read_state NOT IN ('READ','BLOCKED_WITH_REASON') OR disposition_state IN ('UNDECIDED','NOT_APPLIED_YET') LIMIT ?`,\n args:[limit],\n });\n const findings = await sqlite.execute({\n sql: `SELECT id, finding_class, severity, status FROM findings WHERE status NOT IN ('CLOSED','NOT_APPLICABLE') ORDER BY CASE severity WHEN 'CRITICAL' THEN 0 WHEN 'HIGH' THEN 1 WHEN 'MEDIUM' THEN 2 ELSE 3 END LIMIT ?`,\n args:[limit],\n });\n return {artifacts: artifacts.rows ?? [], findings: findings.rows ?? []};\n}\n",[],[168,279],"7fcfac0e-7ad3-48f1-9d1a-71026137e567","reconcile.ts","2592e49e-aa9f-11f1-9eb2-1607ee4eb77e",24,["D",1788773996223],"import { sqlite } from \"https://esm.town/v/std/sqlite/main.ts\";\n\nexport const requiredGates = [\n \"INPUT_ZIPS\",\"INPUT_SET_COMPLETE\",\"FINAL_ARTIFACTS_ALLOWED\",\n \"SEMANTIC_RECONCILIATION\",\"RESEARCH_PACKAGE\",\"AGENTIC_PACKAGE\",\n \"MANUAL_MD\",\"MANUAL_PDF\",\"PDF_EQ_MD\",\"MUTATIONS\",\"CLEAN_ROOM\",\n \"FINAL_RECONCILIATION\",\"FALSIFICATION\",\"RESIDUAL_MATERIAL_FINDINGS\"\n];\n\nexport async function reconcile() {\n const gateRows = await sqlite.execute(`SELECT gate_id,status,numerator,denominator,blocker,updated_at FROM completion_gates`);\n const gates = new Map((gateRows.rows ?? []).map((r:any)=>[String(r.gate_id),r]));\n const reasons:string[]=[];\n for (const id of requiredGates) {\n const g:any=gates.get(id);\n if (!g) reasons.push(`MISSING_GATE:${id}`);\n else if (![\"PASS\",\"TRUE\",\"ZERO\"].includes(String(g.status))) reasons.push(`GATE_NOT_TERMINAL:${id}:${g.status}`);\n }\n const openFindings = await sqlite.execute(`SELECT COUNT(*) AS n FROM findings WHERE status NOT IN ('CLOSED','NOT_APPLICABLE','HISTORICALLY_CLOSED_REVALIDATE_FINAL')`);\n const unresolvedArtifacts = await sqlite.execute(`SELECT COUNT(*) AS n FROM source_artifacts WHERE read_state NOT IN ('READ','READ_COMPLETE','BLOCKED_WITH_REASON') OR disposition_state IN ('UNDECIDED','NOT_APPLIED_YET','REFERENCE_ONLY_PENDING_INPUT_RECONCILIATION','LIBRARY_DERIVED_PARTIAL_PENDING_PHYSICAL_SOURCE_BINDING')`);\n const untestedImpl = await sqlite.execute(`SELECT COUNT(*) AS n FROM implementation_ledger WHERE applied_state='APPLIED' AND test_state!='PASS'`);\n const manualOpen = await sqlite.execute(`SELECT COUNT(*) AS n FROM manual_coverage_requirements WHERE status!='PASS'`);\n const unresolvedSemantic = await sqlite.execute(`SELECT COUNT(*) AS n FROM semantic_objects WHERE status NOT IN ('CANONICAL','MERGED_WITH_PROVENANCE','REFERENCE_ONLY','NOT_APPLICABLE')`);\n const qualityOpen = await sqlite.execute(`SELECT COUNT(*) AS n FROM quality_requirements WHERE status!='PASS'`);\n const counts={\n open_findings:Number(openFindings.rows?.[0]?.n??0),\n unresolved_artifacts:Number(unresolvedArtifacts.rows?.[0]?.n??0),\n applied_but_untested:Number(untestedImpl.rows?.[0]?.n??0),\n manual_requirements_not_pass:Number(manualOpen.rows?.[0]?.n??0),\n unresolved_semantic_objects:Number(unresolvedSemantic.rows?.[0]?.n??0),\n quality_requirements_not_pass:Number(qualityOpen.rows?.[0]?.n??0)\n };\n for (const [k,v] of Object.entries(counts)) if (v>0) reasons.push(`${k.toUpperCase()}:${v}`);\n const open = await sqlite.execute(`SELECT id,finding_class,severity,status FROM findings WHERE status NOT IN ('CLOSED','NOT_APPLICABLE','HISTORICALLY_CLOSED_REVALIDATE_FINAL') ORDER BY CASE severity WHEN 'CRITICAL' THEN 0 WHEN 'HIGH' THEN 1 WHEN 'MEDIUM' THEN 2 ELSE 3 END,id LIMIT 100`);\n return {\n schema:\"grimorio.external-reconciliation-gate/1.1\",\n status:reasons.length?\"FAIL\":\"PASS\",\n reasons,counts,\n required_gates:requiredGates,\n gates:gateRows.rows??[],\n open_findings:open.rows??[],\n invariant:\"NO_TERMINAL_PASS_WITH_UNRESOLVED_SOURCE_FINDING_IMPLEMENTATION_MANUAL_OR_SEMANTIC_STATE\"\n };\n}\n",[],[168,287],["D",1788770443012],[195,196],[168],"b461bbb7-fd78-44ea-92b7-e8c1acba0f3f","V921_SEMANTIC_NONREGRESSION_CONTRACT.md","8edade3e-aa9f-11f1-b321-1607ee4eb77e",20,["D",1788773769830],"# V9.2.1 semantic non-regression contract\n\nRecovered from historical V9.2.1 evidence in File Library. This is a non-regression contract, not proof that current 11-ZIP input contains these exact bytes.\n\n## Known live inventory\n- 258 canonical Skills\n- 15 Skill aliases\n- 60 active Agents\n- 175 active Rules\n- 373 glossary/taxonomy entries\n\n## Universal component classes\nSemantic rigor applies to Skills, Rules, Agents, Validators, Tests, Templates, Routers, Retrievers, Registries, Knowledge Graphs, Scripts, Generators, Runtime Governors, Ledgers, Adapters, Bootstrap, Release Gates, Manifests, Schemas, Config, CI/CD, Migrations, Fixtures, Eval Datasets, Benchmarks, Prompts, System Instructions, Examples, Policies, Crosswalks, Receipts, Reports, Generated/Status Docs, Capability Catalogs, Aliases, Fallback/Dependency Maps, Cleanup/Recovery, Installers, Packaging, Update/Repair/Uninstall, Compatibility Matrices, Run State, State Machines, Concurrency Controls, Tool/Plugin Metadata, Source Provenance and Negative Knowledge stores.\n\n## Core definitions\n- **Operational surface:** any artifact capable of changing what the system selects, executes, believes, validates, writes, delivers or declares as PASS.\n- **Skill:** reusable specialized procedure triggered by intent/phase/risk/gap.\n- **Rule:** invariant/policy restricting system behavior; ENFORCED or ADVISORY with explicit reason.\n- **Agent:** independent-context executor/reviewer justified by isolation, specialization, adversarial review or parallelism.\n- **Validator:** executable detector for a well-defined property; competence must match the claim.\n- **Test:** executable experiment with stimulus, oracle and expected outcome; not synonymous with a structural validator.\n- **Template:** future-artifact mold; headings are not implementation.\n- **Router:** chooses capabilities/branches from intent/context.\n- **Retriever:** retrieves minimum relevant knowledge/authority while controlling provenance/context budget.\n- **Crosswalk:** explicit mapping preserving unmapped/conflict states rather than pretending equivalence.\n- **Disposition:** explicit treatment such as INCORPORATED, MERGED, DOMAIN_ONLY, REFERENCE_ONLY, SUPERSEDED, REJECTED_WITH_REASON.\n- **Operational destination:** where knowledge actually acts; closes SOURCE→CONCEPT→COMPONENT→CONSUMER→VALIDATOR→TEST.\n- **Negative knowledge:** durable failures, rejected hypotheses, unsupported routes and reopen conditions.\n- **Capability Reality Gate:** lifecycle separating declared capability from runtime-verified capability.\n- **Tool Route Receipt:** why a provider/tool was selected/rejected/substituted, including state and limitations.\n- **Generated doc:** derived projection; not an independent source of truth.\n- **Status doc:** current-state projection; must not become stale authority.\n\n## No-lossy operational compression\n`REQUIRED_KNOWLEDGE ⊆ LOCAL_KNOWLEDGE ∪ REFERENCED_KNOWLEDGE ∪ ROUTABLE_KNOWLEDGE`\n\nA smaller component is valid only if all material deeper knowledge remains reachable through validated routing/retrieval. Missing material knowledge becomes REVIEW/UNKNOWN, not model-memory fallback.\n\n## Self-application forward chain\n`SOURCE → CONCEPT/FAILURE/SUCCESS → WHAT IT SHOULD CHANGE → AFFECTED COMPONENT CLASSES → CURRENT IMPLEMENTATION → CONSUMER → ROUTER/TRIGGER → ENFORCEMENT → DETECTOR → TEST → MUTATION → RUNTIME EVIDENCE → STATUS/GAP/FIX`\n\n## Reverse chain\n`ACTIVE COMPONENT → WHY IT EXISTS → SOURCE KNOWLEDGE → CONTRACT → CONSUMER → TRIGGER → OUTPUT → ENFORCEMENT → DETECTOR → EVIDENCE`\n\n## Autonomous operating model\n`DISCOVER → CLASSIFY → ROUTE → RETRIEVE MINIMUM SUFFICIENT KNOWLEDGE → EXECUTE ATOMIC UNIT → OBSERVE → VALIDATE → FALSIFY → RE-ROUTE → CLOSE`\n\n## Release boundary\nA component that exists, is non-empty or is registered is not automatically operational. Operational files require a consumer and contract; derived/reference files require authority/disposition. Runtime behavior requires execution evidence.\n",[],[173,298],"a186b6ac-7b11-4c5e-bb11-826d5abefa8b","STATE_MODEL.md","dea766f6-aa97-11f1-8e67-1607ee4eb77e",5,["D",1788770467721],"# State Model\n\nThe engine tracks evidence without storing raw Grimório text in public project files.\n\n## Artifact lifecycle\n`FOUND -> CENSUSED -> READ -> UNDERSTOOD -> CLASSIFIED -> APPLIED -> TESTED -> CURRENT`\n\nThese states are independent. No downstream state is inferred from an upstream state.\n\n## Completeness gates\n- `ASSURANCE_COMPLETE`\n- `OPERATIONAL_CORE_COMPLETE`\n- `DEEP_CORPUS_INDEX_COMPLETE`\n- `RAW_INPUT_COMPLETE`\n- `MEMORY_COMPILATION_COMPLETE`\n- `MANUAL_KNOWLEDGE_COVERAGE_COMPLETE`\n- `INPUT_SET_COMPLETE`\n- `GLOBAL_HISTORICAL_CORPUS_COMPLETE`\n\n## Core invariants\n- `NO_TEXT_HIT != ABSENCE`\n- `FOUND != READ != APPLIED != TESTED`\n- `SKILL_EXISTS != SKILL_APPLIED`\n- `AGENT_DEFINED != AGENT_EXECUTED`\n- `AUTHENTICATED != AUTHORIZED`\n- `TEST_PASS != CURRENT_EVIDENCE_PASS`\n- `ASSURANCE_COMPLETE != CORPUS_COMPLETE`\n\nRaw source bytes stay outside public code. The project database stores hashes, states, provenance pointers, relationships, findings, validation results, and checkpoints.",[],[173,306],"3f72669e-9f9f-4c00-9cdb-42b6eb3d7c44","NON_REGRESSION_BASELINES.md","56b46b7a-aa9e-11f1-94e9-1607ee4eb77e",14,["D",1788773246127],"# Non-regression baselines\n\nThese are historical evidence floors, not current-input substitutes.\n\n## V3 2026-08-29\n- 3,184 physical manifestations\n- 2,823 CONTENT_READ unique contents\n- 361 DUPLICATED_WITH_REFERENCE\n- 657 deep indexed/routable Skills\n- 16 portable-core Skills\n- 16 logical Agents\n- 24 failure classes\n- JSON 339/339, CSV 84/84, Python AST 429/429, YAML 28/28, JSONL 1/1, TOML 1/1\n- UTF8_DECODE_FAILURE != FILE_CORRUPTION\n\n## GUH V6 2026-08-18\n- 131 canonical Skills\n- 31 specialized Agents\n- 29 always-on Rules\n- 842 logical aliases -> 839 unique objects\n- 30/30 theme coverage\n- 182 Markdown documents crosswalked\n- 7 platform adapters\n\n## GUH V8.1 Living Harness\n- 5,036 current-input manifestations accounted\n- 4,986 text files fully read\n- 50 binary files byte-read\n- 558 physical Skill manifestations crosswalked\n- 242 canonical Skills + 15 aliases\n- 71 Agents\n- 148 Rules\n- 644 compact KB objects\n\n## V9.2.1 reference floor\n- 258 canonical Skills\n- 15 Skill aliases\n- 60 active Agents\n- 175 active Rules\n- progressive disclosure / operational component registry / deep knowledge packs / knowledge reachability\n\n## R5 terminal full-content reference\n- Research subject: 11,013 files\n- Development subject: 11,026 files\n\n## Governing rule\nA smaller current representation is allowed only if it preserves source-manifestation coverage, semantic reachability, provenance, routing/consumer closure, and operational equivalence. Cardinality shrink alone is never proof of regression, but unexplained semantic reachability loss is a release blocker.\n\n`CANONICAL_OBJECT_COUNT != SOURCE_COVERAGE`\n`SEMANTIC_COMPRESSION != PERMITTED_INFORMATION_LOSS`\n",[],[173,314],"bddcb3f5-42a4-4202-9d99-1a099f5d8d18",["D",1788774048394],[],[173,32],"b31636a4-2931-4e92-8a2c-a8da11d24862","CURRENT_PLATFORM_SURFACES_20260907.md","a2f85546-aa9e-11f1-8daf-1607ee4eb77e",15,["D",1788773374068],"# Current platform surfaces — 2026-09-07\n\nThis is currentness metadata for adapter generation. It is not runtime proof on the user's local machine.\n\n## OpenAI Codex\n- Persistent guidance: `AGENTS.md` / `AGENTS.override.md`, hierarchical root -> CWD, closest guidance wins.\n- Repo Skills: `.agents/skills//SKILL.md`; progressive disclosure via metadata then body/resources on use.\n- Skill distribution: plugins; `agents/openai.yaml` may declare invocation policy and MCP dependencies.\n- MCP: external tools/context.\n- Subagents: project/user TOML agent definitions; explicit delegation, own context; thread/depth controls.\n- Rule: keep AGENTS small; move procedures/reference into Skills/docs.\n\n## Anthropic Claude Code\n- `CLAUDE.md`: persistent every-session context.\n- Skills: reusable on-demand knowledge/workflows; Agent Skills-compatible with Claude-specific fields such as `disable-model-invocation`, `context: fork`, agent/background/hooks.\n- Subagents: isolated contexts with tool/permission/MCP/hook/skill controls.\n- Hooks: lifecycle-triggered deterministic automation/enforcement; stronger than prompt-only guidance for invariants.\n- MCP: external service/tool connections.\n- Plugins: packaging layer for skills/hooks/subagents/MCP.\n\n## Gemini CLI\n- `GEMINI.md`: hierarchical project context.\n- Agent Skills: on-demand workflows.\n- Hooks: lifecycle interception.\n- Policy engine: TOML rules with allow/deny/ask_user, tool/MCP/subagent matching and priority.\n- IMPORTANT 2026-09-07: workspace policy tier documented as currently disabled/non-functional; use User/Admin policies until current docs change.\n- Subagents: local/remote, tool/MCP isolation and policy targeting.\n- MCP + A2A supported surfaces.\n\n## GitHub Copilot CLI\n- Instructions: user `~/.copilot/copilot-instructions.md`, `.github/copilot-instructions.md`, modular `*.instructions.md`, AGENTS.md, CLAUDE.md, GEMINI.md.\n- Multiple applicable instruction files are combined; docs do not define a general precedence across all of them. Avoid conflicts; never invent precedence.\n- Skills: `SKILL.md`, auto or slash-invoked, resources/scripts; `allowed-tools` can preapprove tools.\n- Custom agents: Markdown profiles at user/repo/org/enterprise scopes.\n- Hooks: deterministic lifecycle logic.\n- MCP: GitHub MCP preconfigured plus additional servers.\n- Plugins: package agents/skills/hooks/MCP.\n\n## Cursor\n- Rules: `.cursor/rules/*.mdc`; plain `.md` in that directory is ignored by the rule system.\n- `AGENTS.md`: plain Markdown alternative; nested directory support with specific guidance winning.\n- Skills: `.agents/skills/` and `.cursor/skills/`, plus compatibility reads for Claude/Codex skill locations; `SKILL.md` + paths/disable-model-invocation metadata.\n- Subagents: `.cursor/agents/` with compatibility for `.claude/agents/` and `.codex/agents/`; Cursor scope wins naming conflicts.\n- Plugins can bundle rules/skills/subagents/commands/MCP/hooks.\n- MCP supported.\n\n## Devin Desktop / Windsurf lineage\n- Current official Devin docs describe Cascade Skills with progressive disclosure and `.windsurf/skills/`, plus `.agents/skills/` compatibility.\n- Skills != Rules != Workflows:\n - Skill: multi-step procedure + resources; auto or @mention.\n - Rule: behavioral constraint; activation modes.\n - Workflow: manual `/slash-command` prompt template.\n- MCP supported; current branding/product lineage must be checked before emitting final Windsurf/Devin adapter names.\n\n## Google Antigravity 2.0\n- Standalone agent management app, Projects/worktrees, synchronous/asynchronous agents.\n- Tools: system commands, file IO, web, Chrome, Skills, MCP, subagents.\n- Rules/Workflows, scoped settings/permissions and JSON Hooks are distinct surfaces.\n- Artifacts are structured reviewable deliverables (plans, code diffs, diagrams, images, browser recordings).\n- MCP configs vary across 2.0/IDE/CLI/SDK; adapter must target the actual product surface, not a generic name.\n\n## Provider/runtime boundary\nDocumentation currentness != binary installed != authenticated != authorized != quota available != verified for task.\n\nFirecrawl current session: surface connected but search returned HTTP 402 => QUOTA_BLOCKED, not unavailable.\n",[],[173,326],["D",1788770455617],[189,190,191,192,193],[173],["D",1788770368230],["D",1789551045218],{"_98":97,"_106":102},{"_98":97,"_99":100,"_101":102,"_18":53,"_103":104,"_105":24},{"_13":87,"_31":88,"_41":40,"_89":42,"_90":-5,"_91":-5,"_92":-5,"_93":-5,"_94":-5,"_95":43,"_96":97},"ctx",{"_80":-5,"_126":374,"_375":376,"_377":378,"_379":24,"_380":24,"_381":24,"_382":57,"_383":-5,"_384":-7,"_385":-7,"_386":61,"_387":24,"_388":24,"_389":-5,"_390":24,"_391":24,"_392":24,"_393":-5},"ENV",{"_345":346,"_347":348,"_349":88,"_350":351,"_352":353,"_354":355,"_356":357,"_358":359,"_360":361,"_362":363,"_364":365,"_366":367,"_368":369,"_370":371,"_372":373},"REACT_APP_SENTRY_DSN","https://586020d86f05489db63a554690acf1e9@o1398359.ingest.sentry.io/6724499","RENDER_GIT_COMMIT","e87eb04e707cf64e7e532c8e8ec395f209e17f15","RENDER_GIT_BRANCH","POSTHOG_API_TOKEN","phc_nmoyuAkStEZuUp69rYlg1DZgrIt19Gx0Y4Hwf8tGgJP","POSTHOG_API_HOST","https://wart.val.town","WEB_API_TEMPLATE","https://{val}.web.val.run","APP_URL","https://www.val.town","MODULE_URL","https://esm.town","VANITY_WEB_API_TEMPLATE","https://{label}.val.run/","DENOLS_URL","https://lsp.val.town","NODE_ENV","production","CLOUDFLARE_TURNSTILE_SITE_KEY","0x4AAAAAACtaIZJH7s9rBMZ6","GOOGLE_CLIENT_ID","875989696899-jtsqkhlbdv5jrlsk8631p7ipjg3c6o62.apps.googleusercontent.com","STRIPE_PUBLISHABLE_KEY","pk_live_51N2zvWEpZTTLipIkVEwXrUG9AT0yeOOEYxszarZT1jpsZV58Ly6wlg3SltZpHN26CUalOBmUSY7adQRQiSkCZz2500qSXNkf8Y","API_URL","https://api.val.town",[],"preferencesKeybindings","default","preferencesTheme","light","preferencesFixers","preferencesCopilot","preferencesSelectionAi","preferencesCloseBrackets","preferencesChatPanelState","preferencesDprint","preferencesDenofmt","notificationCount","flagDenols","isBusinessOrgMember","businessOrgHandle","isBusinessOrgAdmin","isHigherTownieTier","navCollapsed","currentOrgId"]

Diff · v25