close fullscreen

Administrator+Background Publications Ingestion Plan+Cross-Org Cognitive Repo Sweeps Plan

help edit space_dashboard

Cross-Org Cognitive Repo Sweeps Plan — Code-Extraction Tracker

Working dashboard for the Cross-Org Cognitive Repo Sweeps code-extraction program (cognitive-relevance filter is built into the program name — web / token / marketplace / media repos are out of scope by definition) — the next big extraction phase after the closed PLN / ECAN / OpenPsi / AtomSpace / Perception / AtomSpace-Phase-4 / Non-clustered-HAA / MeTTa-runtime cluster pilots. Snapshot 2026-06-09 (scoping session — split LOCKED, no S1 brief or cloning started yet).

This is a source-code extraction program (GitHub/GitLab org sweeps), distinct from the publications-ingestion backlog tracked in the sibling card Sources Pending Analysis (8160). Filed here because both are “pending-analysis” trackers under the same ingestion-plan parent; do not conflate code-extraction with publications-ingestion.

Standing protocol (carry forward to every subcluster)

  • Scoped-Gemini Option (i): Codex is citation-grade (file:line cites, fork/branch counts, quirk-applicability); Gemini is synthesis/architecture/identity/placement only. Relay both with explicit “review only — do not write files or wiki.” Orchestrator (Claude) is the one wiki writer; verify after every write.
  • Cognitive-relevance filter (LOCKED this session): for blockchain/marketplace-heavy orgs, pre-filter each org’s repo list to AGI / Hyperon / MeTTa / AtomSpace / cognitive-architecture repos only. Log+exclude pure token/marketplace/ledger/Ocean-data/frontend plumbing in the brief (don’t silently drop — record what was excluded and why).
  • No-re-extract discipline: several target orgs have repos already covered by prior pilots. Sweep residual only; cite the prior pilot for the covered repo.
  • Git: pull-before-extract per source; record HEAD SHA in each brief. No cloning until a subcluster’s split is locked and its S1 brief is authored.

Subcluster split (LOCKED 2026-06-09)

# Subcluster Org(s) / scope Inherited / excluded
SC-1 ASI-Alliance Core (cognitive) github.com/asi-alliance, cognitive-filtered exclude token / marketplace / Ocean-data plumbing
SC-2 Agent Frameworks github.com/fetchai (uAgents / CosmPy / agentverse) + asi agent SDKs scope to agents-as-Hyperon-consumers, not framework internals
SC-3 Distributed Compute gitlab.com/nunet + CUDOS-under-asi genuinely-new substrate axis (no prior pilot)
SC-4 Decentralized Runtime / F1R3FLY github.com/F1R3FLY-io residual MeTTaTron done (MeTTa-runtime S3); f1r3node / Rholang / RChain remain
SC-5 Bio / Longevity github.com/Rejuve* residual neural-subgraph-matcher-miner done (Perception pilot); COMPLETE — S1 query-core + S2 agent-layer + S3 experiments CLOSED 2026-06-21 (applied Hyperon consumer)
SC-6 iCog-Labs github.com/iCog-Labs-Dev residual conceptBlending / metta-moses done (HAA S4/S5)
SC-7 Mindplex / Xccelerando github.com/Xcceleran-do (= Mindplex org) cognitive subset only; exclude web/media/token

Excluded from scope this session: Magi-AGI (our own org — Lake’s call; overlaps live implementation/ingestion workstreams). hansonrobotics/* is substantively covered by OpenPsi (2026-04-28) + Perception (2026-05-01); only residual non-OpenPsi/non-perception Hanson repos would need a separate sweep, not part of this program.

Locked pilot ordering (smallest / best-bounded → largest)

SC-4 F1R3FLY-residual → SC-5 Bio → SC-7 Mindplex → SC-3 Distributed-Compute → SC-6 iCog → SC-2 Agent-Frameworks → SC-1 ASI-Core.

Rationale: ride the inheritance (SC-4 inherits MeTTaTron, SC-5 inherits the miner), build the cognitive-relevance-filter discipline on small bounded orgs before hitting the giant relevance-filter-dependent SC-1 ASI-Core last.

Per-subcluster source inventory

Repo lists below are to be verified at each subcluster’s S1 brief (fresh org listing + git pull); only SC-7 was repo-enumerated this session. Do not treat unverified repo names as ground truth.

  • SC-7 Mindplex / Xccelerando — VERIFIED 2026-06-09 (WebFetch github.com/Xcceleran-do). Cognitive subset: OmegaClaw-Core (Python, forked from asi-alliance), mettaclaw (MeTTa, forked from patham9), Mindplex-Hyperon (Python Hyperon impl), snet-aigents-service (Python AI agents), mindplex_semantic (TS vector search / conversational AI). Excluded (web/media/infra): mindplex_landing, Mindplex-landing-page, mindplex_docs, mindplex-keycloak-theme, mindplex-chat-backend, mindplex-token, ip_location_lookup_api, image-compression-service.
  • SC-4 F1R3FLY-residual — INVENTORIED + SPLIT LOCKED 2026-06-09 (S1 brief authored; scripts/archive/cross_org_sweeps/sc4_f1r3fly_residual/). Fresh org listing = 57 repos. Cognitive residual after MeTTaTron (S3) + Scala MeTTaIL (S4) excluded. 3-source split: S1 = mettail-rust (Rust language! framework, HEAD a95726a 2026-05-20, 299 commits ACTIVE) + mettail-rust-zariuq (Zar’s Lean-export fork) ← CLOSED 2026-06-09 (Codex citation-grade + Gemini synthesis-only + orchestrator spot-verify). Verdict: [LANGUAGE-SPECIFICATION-FRAMEWORK] / [RUST-NATIVE-PARALLEL-SUCCESSOR-CLEAN-SHEET] of Scala MeTTaIL (related by GSLT vision, NOT code) / [PARSER-PRATTAIL] (LALRPOP removed; tree-sitter is zariuq-branch-only) / [ASCENT-REWRITE-ENGINE] / [PATHMAP-WIRED-AS-RHOCALC-LITERAL-OPS] (real pathmap dep, scoped — not a MORK merge) / [ROCQ-PROOF-SUITES-REAL] (12 .v, 0 Admitted/124 Qed) / [IL-TRACK-FORMAL-SEMANTICS-NOT-RUNTIME]. 2 wiki writes (7268 + 4632, mettail-rust added beside Scala MeTTaIL). Gemini 6/12 aligned (Codex corrected parser/pathmap-scope/.v-stubs/identity). S2 (OSLF + hypercube + OSLF-editor + rho2riscV + graphl-parser) CLOSED 2026-06-21 — verdict [OPERATIONAL-SEMANTICS-AS-DATA-TOOLING-TRACK] (not a runtime, no working compiler at HEAD): OSLF [DESIGN-STUB] / hypercube [FORMAL-SEMANTICS-SPEC-PARSER] (parses .gslt, doesn’t execute it) / OSLF-editor [TOOLING-INFRA] Blockly editor / rho2riscV [COMPILER-SPEC-WITH-REFERENCE-IMPL] (rho-machine levels 0-1, no RISC-V emitter) / graphl-parser [TOOLING-INFRA] (BNFC GraphL parser; stale “Graph to RhoLang” label rejected). GSLT parsed-not-executed outside mettail-rust (S1); GSLT2rho branch-only. Gemini 100% aligned with Codex (cleanest Option-(i) turn). 1 grouped wiki write (7268 OSLF/GSLT satellites). Scope: graphl-parser + OSLF-editor de-scoped from later SC-4; OSLF no deeper pass. Archives in scripts/archive/cross_org_sweeps/sc4_f1r3fly_residual/ (S1+S2, _S1/_S2 names; externally backed up after a parallel-session force-add/branch-switch swept the originals — recovered from commit 235cbaa). S3 (rhoHDC + Hypervector + itm + MacroLoop) CLOSED 2026-06-21 — SC-4 COMPLETE (3/3). S3 verdict [ERA-A-SCALA-RESEARCH-PROTOTYPE-LAYER]: rhoHDC [NEURAL-SYMBOLIC-HDC-ENCODER] → Hypervector [HDC-VSA-LIBRARY] → MacroLoop [METAPROGRAMMING-LIBRARY-FORK-MIRROR] (real Scala dep chain); itm [TRIE-MAP-RESEARCH-PROTOTYPE] = conceptual cousin of MORK/PathMap, NO code dep. MORK is NOT a code dependency of any SC-4 repo (pathmap-dep in mettail-rust + Vandervorst identity-lineage only). Gemini 100% aligned (2nd clean turn). 1 wiki write (7268 Era-A group). SC-4 subcluster-final synthesis in findings_reconciled_crossmodel_S3.txt: F1R3FLY residual = formal-semantics/language-framework research line (not a runtime), Era-A Scala (OSLF/GSLT + HDC/trie/metaprog) → Era-B Rust (mettail-rust, only production-grade artifact + Rocq proofs). De-scoped: all S3 repos + graphl-parser + OSLF-editor. S2 = OSLF + hypercube + OSLF-editor + rho2riscV + graphl-parser (operational-semantics / GSLT / compilation); S3 = rhoHDC + Hypervector + itm + MacroLoop (Era-A Scala research: rho↔HDC, trie maps, Vandervorst metaprogramming). Substrate OUT (f1r3node / rholang / rholang-rs = blockchain/concurrency infra; cognitive payload is the MeTTaIL layer that compiles TO rho). Pubs-track cross-links (not code extraction): MeTTaCycleYellowPaper + F1R3FLY publications repo. Two-era narrative: Era-A Scala prototypes (2023-25) → Era-B Rust framework (2026). Key open Q for S1: does mettail-rust supersede S4 Scala MeTTaIL / relate to S3 MeTTaTron? PathMap wired into runtime/ at HEAD (pathmap_bridge/codec/lit.rs) — cross-link MORK family map 8441.
  • SC-5 Bio — INVENTORIED 2026-06-21 (rejuve-bio 39 cognitive subset + Rejuve.AI 10 repos OUT); S1 query-core CLOSED 2026-06-21 (flagship annotation-query-backend + 4 satellites). neural-subgraph-matcher-miner DONE — exclude.
  • SC-3 Distributed-Compute — to-inventory: gitlab.com/nunetgroups/repos; CUDOS repos under asi-alliance.
  • SC-6 iCog — to-inventory: iCog-Labs-Dev residual. conceptBlending + metta-moses DONE — exclude.
  • SC-2 Agent-Frameworks — to-inventory: fetchai uAgents / CosmPy / agentverse; asi agent-marketplace SDKs.
  • SC-1 ASI-Core — to-inventory: asi-alliance org under cognitive whitelist (resolve at S1).

Cross-pilot identity / cross-link hooks

  • OmegaClaw-Core (SC-7) ↔ the “Omegaclaw PLN” Mattermost publications family (~5 papers) in tracker 8160. Code↔publications cross-link when both are synthesized.
  • mettaclaw (SC-7) forked from patham9 — the PeTTa / metta-nars / NACE author across MeTTa-runtime (S2) + HAA (S5) pilots. Cross-pilot identity already mapped.
  • Cross-pilot identity map (carry forward): Stay/Meredith (F1R3FLY), Treutlein (PeTTa), Miles=logicmoo=TeamSPoon (MeTTaLog), Khudobakhshov+Potapov (jetta), Vepstas (atomspace-metta / AtomSpace-Scheme), zariuq=Zar Goertzel, ngeiswei=Geisweiller.

Open bounding questions (resolve at each subcluster’s S1 — non-blocking for the split)

  1. CUDOS — placed in SC-3 Distributed-Compute (CUDOS joined the ASI Alliance as the compute layer); confirm at SC-3 S1.
  2. Ocean Protocol — data-marketplace tooling → excluded by cognitive filter; confirm no cognitive-architecture repos slip through at SC-1.
  3. asi-alliance cognitive whitelist — the exact repo set for SC-1 needs a full org listing + per-repo relevance call at SC-1 S1.
  4. SC-4 F1R3FLY residual boundary — RESOLVED 2026-06-09: cognitive residual = mettail-rust(+zariuq) / OSLF / hypercube / OSLF-editor / rho2riscV / graphl-parser / rhoHDC / Hypervector / itm / MacroLoop. Substrate (f1r3node / rholang / rholang-rs) ruled OUT. See SC-4 row above.

Decisions locked this session (2026-06-09)

  • Scope filter on big orgs = cognitive-relevance filter (not full-sweep-tag-and-defer).
  • Magi-internal = out of scope; Xcceleran-do = in scope (identity confirmed = Mindplex / Xccelerando Media, Goertzel Chief Scientist; has a real ~5-repo cognitive subset).
  • Ordering = accepted as proposed (SC-4 first, SC-1 last).
  • Card home = here, under Administrator+Background Publications Ingestion Plan, beside tracker 8160.

Next step

SC-4 (F1R3FLY-io residual) COMPLETE — all 3 sources CLOSED (S1 mettail-rust+zariuq 2026-06-09; S2 OSLF/GSLT satellites + S3 Era-A Scala prototypes 2026-06-21 — see SC-4 row above). Subcluster-final synthesis archived. SC-5 Bio Sources 1 + 2 CLOSED 2026-06-21 (split: S1 query-core / S2 MeTTa-LLM agents / S3 neuro-symbolic). S1 [APPLIED-BIOMEDICAL-HYPERON-QUERY-CORE] = applied Hyperon consumer (flagship embeds MeTTa + clients MORK, not a runtime/fork). S2 [APPLIED-METTA-LLM-AGENT-LAYER]: metta-motto-biochatter (zarqa-ai/metta-motto fork + bio-RAG); bio-semantic-parser (active divergent fork of iCog, NLP→relations); BioClaw (multi-agent biocurator on OmegaClaw, code-backed PLN Truth_Revision over MeTTa + MORK-staged HITL); biochatter-metta = iCog CI-mirror→SC-6; AI-Assistant/chatbot-ALIS = non-Hyperon apps. 4 strong Option-(i) turns; 2 wiki writes to 7268. S3 [BIO-NEURO-SYMBOLIC-EXPERIMENTS-ONE-HYPERON-TWO-NON-HYPERON]: galaxy-htn-neurosym (partial Hyperon HTN/PLN/MeTTa scaffold — planner/reasoner empty stubs, mock STVs); hypothesis-generation-demo (Prolog/LLM, non-Hyperon); galaxy-link-pred (graph-ML, non-Hyperon). SC-5 (Bio/Longevity) COMPLETE — all 3 sources CLOSED 2026-06-21. Subcluster thesis: rejuve-bio = applied biomedical Hyperon CONSUMER (embeds MeTTa, clients MORK, builds on metta-motto+OmegaClaw, genuine PLN Truth_Revision in BioClaw), not a runtime/framework; cross-org deps iCog (SC-6) + Mindplex/OmegaClaw (SC-7). 5 strong Option-(i) turns; 3 wiki writes to 7268. SC-7 Mindplex COMPLETE 2/2 (2026-06-23). S1 (Mindplex-Hyperon) CLOSED 2026-06-22 — verdict [APPLIED-HYPERON-PETTA-RECOMMENDER-MINING-STACK]. Mindplex-Hyperon is an applied Hyperon MeTTa/PeTTa consumer (deps hyperon pip + PeTTa submodule trueagi-io/PeTTa; MIND-benchmark ingestion → native-MeTTa frequent-pattern-miner (ETV) → PeTTa backward-chainer → Flask mining/sim API; standalone MeTTa PLN ruleset not on active path), NOT “classical OpenCog AtomSpace” (README rhetoric; zero opencog imports). Active surface demo-PeTTa (pin 8fbd847); main stale doc-only (9c5468f); Neuro-Symbolic (1cda1db) mixed (mettaMotto Hyperon / DFOL non-Hyperon). Codex refinement adopted: MORK/PathMap are dangling .gitignore’d gitlinks — NOT app-used (brief’s “applied MORK use” assumption refuted; not cross-linked to 8441). End-to-end partial-but-working (no scaffold stubs). 6th consecutive strong Option-(i) turn; 1 wiki write to 7268. SC-7 S2 (Claw family) CLOSED 2026-06-23 — [HYPERON-PETTA-CLAW-AGENT-RUNTIME-FAMILY]: patham9/mettaclaw (canonical lean MeTTa-native core, Hammer NAL-1–7, ChromaDB AtomSpace LTM) + asi-alliance/OmegaClaw-Core (productized neural-symbolic framework, agent-local PLN lib_pln, Telegram/Docker). Lineage = common-root git fork of mettaclaw (shared root afbfe9c; diverged merge-base 1a7c01e 2026-04-10; 614 ahead + bidirectional cross-pollination). Mindplex forks stale (mettaclaw 0/53; OmegaClaw 4/265 = AWS Bedrock provider delta only, no new cognition). openclaw/vericlaw = different JS/TS upstream (openclaw/openclaw, godelclaw/vericlaw). BioClaw (SC-5) builds on OmegaClaw; OmegaClaw home = SC-1 asi-alliance — extracted here, cross-ref not re-extract at SC-1. 7th consecutive verdict-aligned turn (2 Gemini process flags: Option-(i) cite-supply breach + premature extraction; both acknowledged). 1 wiki write to 7268. SC-3 Distributed Compute SCOPED 2026-06-23 — shape LOCKED = 1 HERMES-centered source. Org listings: NuNet (gitlab.com/nunet)SCOPEUnderHERMES 133 projects; CUDOS (CudoVentures) 60 repos. CRITERION EXPANDED (Lake 2026-06-23, program-wide): also include non-Hyperon-origin tech that AUGMENTS Hyperon workflows (so teams can absorb it), not just Hyperon-native code; still exclude general-AI-beyond-stack + blockchain/infra plumbing. it: CUDOS = exclude entirely (pure Cosmos blockchain/cloud-compute). NuNet = mostly infra, but (LLM-agent platform: CLI/messaging/TUI agent core, plugins/skills, MCP, cross-session memory) is IN as community-adopted augmentation-workflow tech — NON-Hyperon-origin (0 hyperon/metta hits; NOT the Magi HERMES card 741, name coincidence). SC-3 S1 = deep HERMES workflow doc (absorb-this artifact) + brief NuNet augmentation map (MCP servers/agent-archetypes/ml-on-nunet) + boundary-exclude blockchain/infra; f1r3fly-node → SC-4 cross-link. Likely a dedicated workflow card. SC-3 S1 BRIEF-REVIEW COMPLETE 2026-06-23: canonical HERMES pinned (NousResearch/hermes-agent@807bdc17 primary + nunet/hermes-agent-lite@7c2c72a1 adaptation; lite delta = 7 ahead / 298 behind, merge-base 5aec00f7, entire delta is thin-image packaging). S1 brief authored + relay packets inlined; Codex signed off after 5 pre-dispatch edits (all applied: §0 local paths; Q9 divergence method+precompute; 807bdc17 IST commit-date; Q10 scoped-grep+false-positive classification; Q9 strip-set diff/Dockerfile.lite cites); Gemini signed off clean. SC-3 CLOSED 2026-06-23 (1/1 + subcluster-final): both models extracted @ pins + verdict-aligned [NON-HYPERON-SELF-IMPROVING-AGENT-WORKFLOW-TO-ABSORB]; orchestrator spot-verified all load-bearing cites @ pins (Q10 zero-Hyperon scoped grep = 0 source hits; Q9 7/298 delta = thin-image only, uv-sync drops slack-extra; Q5 prompt-cache AGENTS.md:19 + prompt_caching.py:49; Q4 MCP client+server FastMCP; Q3 memory core + external provider plugins). Wiki writes: NEW top-level External Agent Workflows (9538) parent stub + External Agent Workflows+HERMES Agent Workflow (9541) dedicated card (both Draft, ai_generated; placement = standalone top-level per Lake lean + Codex — boundary clarity, NOT under Implementation Families) + 7268 SC-3 entry. SC-6 iCog-Labs SCOPED 2026-06-23 (shape LOCKED = 3 sources, Option A): iCog-Labs-Dev = 80 repos, mostly already-done by prior pilots (MOSES/MetaMo/ConceptBlending/ECAN/miner/MORK). Residual split: S1 bio carry-in (biochatter-metta lib+server+client [2024-era] + bio-semantic-parser [active 2025-11] + pubchem2metta + bio-data-semantic-parsing); S2 new iCog-native cognitive (SubRep ★ certificate-driven skill discovery + Qwestor-PeTTa + Pattern-miner-mm2 + PeTTa-OpenPSI); S3 augmentation tech (MeTTa-AI-Assistant RAG-for-MeTTa + mcp-xp MCP+Galaxy + galaxy-agent-xp-II) + boundary note (neural/predictive-coding tier + infra + already-done cross-refs; cordial-f1r3node→SC-4). S1 CLOSED 2026-06-23 (bio carry-in): verdict [APPLIED-METTA-BIOMEDICAL-NL-KG-CONSUMER-STACK]; both models extracted + verdict-aligned 10/11. Load-bearing dissent resolved at pin: biochatter-metta IS a real Hyperon runtime consumer (hyperon==0.1.6, MeTTa().run()) — Gemini’s blanket [EMITTER-NOT-INPROCESS-RUNNER] REJECTED; reconciled to per-repo [MIXED-COUPLING] (biochatter-metta runs MeTTa; bio-semantic-parser+pubchem2metta emit-only). iCog=canonical origin, rejuve-bio downstream (bio-semantic-parser 37/0, biochatter-metta 12/0 CI-only). Wiki: 7268 grouped entry (no dedicated card). S2 READY TO RELAY (2026-06-23): 4 repos cloned/pinned (subrep@d128ba91 ACTIVE 94.py+1.metta cds-pds/skill-gen/MDN branches; Qwestor-PeTTa@7f2b701c 21.metta; Pattern-miner-mm2@d95b577c 3.metta surp branch; PeTTa-OpenPSI@f59d3c5c 94.metta nars/minecraft branches). S2 brief authored (individual per-repo verdicts + fail-closed on subrep CDS/PDS + hyperon-bridge wired-vs-stub + SubRep paper-alignment to Goertzel 2025-09/HAA-S1) + relay packets built. S2 CLOSED 2026-06-23 (new iCog-native, 4 distinct lines): [ICOG-NATIVE-MIXED-RESEARCH-LINES-SUBREP-IMPLEMENTATION-BACKED]. subrep [SUBREP-CERTIFIED-SKILL-DISCOVERY-HYPERON-CERTIFICATE-STORE] (impl-backed CDS/PDS gates + REAL hyperon cert store ≥0.2.0; closes HAA S1 SubRep narrative guard; paper-aligned NOT AIRIS-derived); Qwestor-PeTTa [PETTA-LLM-CONTEXT-ACTION-ROUTER] (Gemini-2.5 context→MetaMo routing; NOT query/KG); Pattern-miner-mm2 [MM2-SURPRISINGNESS-MINER-PARTIAL] (surp.metta only, frequent-miner empty; ≠HAA-S3 miner); PeTTa-OpenPSI [PETTA-HYPERON-OPENPSI-PORT-WITH-PLN-NARS] (hyperon==0.2.3+Python hooks; 3rd OpenPsi artifact). TWO load-bearing Gemini drifts rejected at pins (subrep coupling = wired not stub [self-corrected]; Qwestor = LLM-router not query-reasoner [NOT self-corrected]). Wiki: 7268 grouped entry + SubRep Deep Dive card created at Publication Maps+SubRep Deep Dive (9639, Draft, ai_generated) — standalone, OUTSIDE the curated topic Index per Lake (Index/IndexSubtopic cards are off-limits; grouped with the MORK Research Family map; cross-links AIRIS, not nested under it). S3 READY TO RELAY (2026-06-23, FINAL SC-6 source): 3 repos cloned/pinned (MeTTa-AI-Assistant@c85f077f origin/DEV 576c/96.py RAG-for-MeTTa; mcp-xp@b63d7d49 custom MCP server+multi-LLM/Galaxy; galaxy-agent-xp-II@086c5c52 LLM workflow recommender SBERT+Gemini). S3 brief authored (individual per-repo verdicts + fail-closed + per-repo Hyperon/MeTTa coupling discipline [S2 lesson] + IN/OUT boundary + SC-6 subcluster-final synthesis) + relay packets built. S3 CLOSED 2026-06-23 — SC-6 COMPLETE (3/3). S3 verdict [ICOG-AUGMENTATION-TECH-METTA-RAG-MCP-GALAXY-TOOLS]: MeTTa-AI-Assistant [METTA-RAG-DEV-ASSISTANT-WITH-FRONTEND-HYPERON-WASM-PLAYGROUND] (MIXED — RAG-over-MeTTa-text backend + real Hyperon v0.2.10 WASM frontend playground; Gemini’s text-only verdict REJECTED at pin); mcp-xp [MCP-GALAXY-MULTI-LLM-BIOBLEND-TOOL-BRIDGE-ZERO-HYPERON]; galaxy-agent-xp-II [GALAXY-LLM-EMBEDDING-RECOMMENDER-ZERO-HYPERON]. SC-6 subcluster portrait [ICOG-LABS-RESIDUAL-APPLIED-CONSUMER-PLUS-NEW-NATIVE-PLUS-AUGMENTATION]; load-bearing guard = per-repo coupling (in-process Hyperon ↔ MeTTa-text-only ↔ zero-Hyperon). 4 Gemini coupling/identity drifts across SC-6’s 3 sources, all caught by Codex+orchestrator. Wiki: 7268 S3 group + subcluster summary. SC-2 Agent Frameworks SCOPED 2026-06-23 (shape LOCKED = 2 sources, Option A): fetchai = 96 repos, mostly blockchain/ledger/Cosmos infra (OUT). Augmentation surface = uAgents agent framework + ASI-Alliance ecosystem. S1 uAgents framework (decorator agents + Almanac on-chain discovery + protocols + crypto + experimental LLM chat-agent) + agentverse-skills (FIRST SKILL.md skills for the ASI ecosystem — ties to HERMES SC-3) + ecosystem map (fetchai/ai-engine/langchain-uagents/AEVS) + AEA-legacy note. S2 the Hyperon-coupling bridge: innovation-lab-examples/web3/singularity-net-metta* (uAgent↔SingularityNET-MeTTa RAG agents — the genuine agents-as-Hyperon-consumers) + A2A/MCP examples. uAgents core zero-hyperon; singularity-net-metta = 15 metta/13 hyperon hits. S1 repos cloned/pinned (uAgents@c210bbea + agentverse-skills@16fab531 + 4 ecosystem) + S1 brief authored + packets built. S1 CLOSED 2026-06-24 (uAgents framework + agentverse-skills): [EXTERNAL-AGENT-FRAMEWORK-AND-SKILL-WORKFLOW-TO-ABSORB]; FIRST fully verdict-aligned source (no load-bearing dissent). uAgents [DECENTRALIZED-UAGENTS-FRAMEWORK-WITH-ALMANAC-BOUNDARY] (decorator async agents + Almanac CosmWasm registration/discovery + typed protocols + ChatAgent/ASI:One + uagents-adapter w/ MCP/A2A/LangChain/CrewAI); agentverse-skills [PORTABLE-SKILLMD-AGENTVERSE-WORKFLOW] (8 SKILL.md skills, anthropics/skill-md-spec — SAME standard as HERMES SC-3); ecosystem map (fetchai legacy / ai-engine / langchain-uagents / AEVS audit). All 6 repos ZERO-hyperon (orchestrator-verified all file types). Wiki: 7268 S1 group. S2 READY (2026-06-24, FINAL SC-2 source): innovation-lab-examples@71f35a5b cloned/pinned + S2 brief authored + packets built. GENUINE Hyperon coupling CONFIRMED at pin: 3 web3/singularity-net-metta* examples = uAgent fronting an IN-PROCESS Hyperon MeTTa knowledge space (from hyperon import MeTTa+MeTTa()+metta.space().add_atom(E(...))+metta.run(); hyperon>=0.2.6) for medical/general/investment RAG + ASI:One LLM; A2A/MCP examples zero-Hyperon (all 13 hyperon hits in the 3 bridge dirs, 0 outside; no AtomSpace/PLN/MORK/PeTTa). Codex extracted [UAGENTS-HYPERON-METTA-BRIDGE-EXAMPLES-PLUS-A2A-MCP-INTEGRATION] (trigger arrived before brief = Lake state-mixup, not a process issue; orchestrator-confirmed). S2 CLOSED 2026-06-24 — SC-2 COMPLETE (2/2). S2 [UAGENTS-HYPERON-METTA-BRIDGE-EXAMPLES-PLUS-A2A-MCP-INTEGRATION]: 3 web3/singularity-net-metta* examples = uAgent fronting an in-process Hyperon MeTTa space (medical/general/investment RAG + ASI:One); A2A/MCP examples zero-Hyperon; coupling narrow (orchestrator-confirmed). Both models verdict-aligned (2nd clean source). SC-2 final portrait [FETCHAI-AGENT-FRAMEWORKS-WITH-NARROW-HYPERON-BRIDGE-EXAMPLES]: Fetch.ai=NEW(theSC-1KEYS1 ASI-Alliance decentralized agent-network infra (S1, zero-Hyperon) + applied coupler (S2 narrow bridge); MCP thread HERMES→mcp-xp→uAgents. Wiki: 7268 S2 group + subcluster summary + dedicated card External Agent Workflows+uAgents Agent Framework (9642, Draft, ai_generated) beside HERMES 9538 parent had reserved the fetchai/SC-2 slot). ASI-Alliance Core SCOPED 2026-06-24 (shape LOCKED = 1 source + program close, Option A) — FINAL subcluster. FINDING: asi-alliance org = 16 repos, THIN on new cognitive code — mostly ASI-Chain blockchain (OUT); cognitive core is elsewhere (OmegaClaw=SC-7 done; Hyperon=trueagi-io/singnet original pilots). = extract import-knowledge-package (agent-LTM utility: distilled knowledge JSONL/MeTTa → ChromaDB vector embeddings via OpenAI/SentenceTransformers; /PeTTa/chroma_db; pre-read zero-Hyperon-runtime) + SURVEY OmegaClaw deployments (Oma_folio crab-agent + Max_folio MeTTaClaw-agent artifacts = SC-7-Claw-in-production; cross-ref) + boundary (ASI-Chain OUT, asi-chain-node→SC-4, max-music general-AI OUT, ASI:One/Agentverse→SC-2) + PROGRAM-CLOSE synthesis. S1 repos cloned/pinned (import-knowledge-package@c82268ec + Oma_folio@211a6203 + Max_folio@104ceb65) + S1 brief authored + packets built. SC-1 CLOSED 2026-06-24 — CROSS-ORG COGNITIVE REPO SWEEPS PROGRAM COMPLETE (7/7 subclusters). SC-1 [ASI-THIN-CLOSE-AGENT-LTM-UTILITY-PLUS-CLAW-ARTIFACTS]: import-knowledge-package = ChromaDB agent-LTM importer (MeTTa-as-data, zero Hyperon runtime; KB-data mentions ≠ code coupling — the program’s coupling discipline in its purest closing form); Oma_folio/Max_folio = SC-7 OmegaClaw agent-output artifacts (survey, cross-ref); ASI-Chain blockchain + max-music art OUT. 3rd consecutive fully verdict-aligned source. ALL 7 SUBCLUSTERS CLOSED: SC-4 F1R3FLY + SC-5 Bio + SC-7 Mindplex + SC-3 Distributed-Compute + SC-6 iCog + SC-2 Agent-Frameworks + SC-1 ASI-Core. Program portrait [CROSS-ORG-SWEEPS-COMPLETE-7/7]: adjacent orgs built a modular applied-AI ecosystem AROUND Hyperon (not re-writing the interpreter) via 5 threads (LLM-to-MeTTa / MeTTa-native ports / MCP-A2A axis HERMES→mcp-xp→uAgents / SKILL.md standard / the Claw agent line); standout new-native = SubRep (iCog SC-6); load-bearing discipline = per-repo commit-level coupling verification. Wiki: 7268 SC-1 group + subcluster summary + program-close synthesis; new External Agent Workflows tree (HERMES 9541 + uAgents 9642) + Publication Maps+SubRep Deep Dive (9639). PROGRAM COMPLETE — no further subclusters. Per locked ordering (SC-5 Bio → SC-7 Mindplex → SC-3 Distributed-Compute → SC-6 iCog → SC-2 Agent-Frameworks → SC-1 ASI-Core). SC-5 = github.com/Rejuve* residual; neural-subgraph-matcher-miner already done (Perception pilot) — exclude. S1 prep (org listing, clone, pin, git-history) is done; S2 prep = clone metta-motto-biochatter / biochatter-metta / bio-semantic-parser (+ AI-Assistant / BioClaw / chatbot-ALIS), pin, and resolve the biochatter-metta vs iCog-Labs-Dev/biochatter-metta lineage (decision B3) at extraction. Carry-in: scoped-Gemini Option (i) is the stable baseline (2 consecutive 100%-aligned turns at SC-4 S2/S3). This card is the authoritative dashboard; update per-subcluster as each S1 opens / closes.