allaun/Research-Stack
1
0
Fork 0
mirror of https://github.com/allaunthefox/Research-Stack.git synced 2026-07-31 03:05:21 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
757 commits 18 branches 0 tags 425 MiB
Commit graph

25 commits

Author SHA1 Message Date
allaun
2cbba82406 feat(agent): break-glass MCP agent for novel problem attacks 
Last-resort multi-model panel via OpenRouter Fusion. Only invoked
when all other approaches (local model, single frontier, existing
skills) have failed on the same problem.

Architecture:
- Problem classification (lean_proof, code_bug, architecture, math_novel, infra_debug)
- Per-type model panel selection (Opus + DeepSeek + Gemini for proofs)
- Context assembly (files, build logs, AGENTS.md constraints)
- OpenRouter Fusion call (parallel panel + judge synthesis)
- Structured output (diagnosis, solution, alternatives, risks, cost)

Safety:
- Explicit 'break glass:' trigger required
- Cost gate: /usr/bin/bash.30-.50 per invocation, .00/session cap
- Rate limit: 3 invocations per session
- Audit trail via ContextStream

Files:
- 4-Infrastructure/shim/break_glass_mcp.py (MCP server)
- 6-Documentation/docs/specs/break_glass_mcp_agent.md (design spec)
- .opencode/skills/break-glass/SKILL.md (skill definition)
- .mcp.json (server registration)
 2026-06-22 15:33:16 -05:00
allaun
7098ee0d64 docs(specs): add SilverSight specification draft with Core Semantics sections 
Adds 6-Documentation/docs/specs/SilverSight_Spec.md, documenting the
YaFF-inspired separation of schema, layout, access pattern, wire format,
view, layout bridge, and canal-aware layout selection for the SilverSight
clean-slate architecture.
 2026-06-22 01:11:17 -05:00
allaun
b6f088d9f1 docs(rrc): add topological braid adapter spec for fibonacci anyons 
Define the mathematical isomorphism between the existing compiler concepts (ColorRope, stairIndex, tensegrityCoherent, phi_pow, Zeckendorf bits) and the Fibonacci anyon braiding/fusion algebraic coordinates.

Build: 0 jobs, 0 errors (lake build - no code changes)
 2026-06-18 22:28:50 -05:00
allaun
558431fc45 feat(lean,infra): Corpus278→250 rename + SLOS-calibrated braid defaults 
- Renamed Corpus278/Matrices278 to Corpus250/Matrices250 across all
  Lean modules, Python shims, JSON schemas, and documentation
- Applied SLOS-calibrated defaults to braid_search.py:
  bracket_cost base=8x/16x, crossing_penalty gap_reward=0.01
- Synced same defaults to research-compute-fabric submodule
- Updated Burgers 0D defaults: nu=0.9995, advection=0.075
- Fixed 278→250 refs in pist_predictions JSON (claim_boundary, total)
- Fixed stale schema refs in .devin/skills/lean-proof/SKILL.md
- Updated AGENTS.md build baselines to post-clean counts (3314 jobs)
- Updated lake-manifest.json sparkle rev to match lakefile.toml
- Fixed Q16_16 signed conversion bug in qaoa_adapter tunable costs
- NBody.lean: pre-existing dirty changes (introN failures, NOT our work)

Build: 3314 jobs, 0 errors (lake build Compiler)
 2026-06-18 22:16:52 -05:00
allaun
77488ac0ae feat(lean): close gaussian_line_integral_unit_dir + consolidate infrastructure 
Lean proof fixes:
- N3L_Energy.lean: fully close gaussian_line_integral_unit_dir (nlinarith+hab
  for unit-circle quadratic, sqrt_mul+neg_div for integral_gaussian_1d match,
  exp_sum_of_sq order fix, add_assoc for h_gauss_shift, sq_sqrt for field_simp,
  sq_abs for perpDistance hd)
- Add Adapters/AlphaProofNexus: 12 Erdos/graph adapter stubs (AlphaProof nexus)
- Add Adapters/ErgodicAdditive.lean, SidonMatroid.lean
- Add AntiDiophantine.lean, EffectiveBoundDQ.lean, PVGS_DQ_Bridge.lean
- Add FormalConjectures/Util/ProblemImports.lean
- Add RRC/EntropyCandidates/Candidates.lean
- Add OTOM external project (lakefile.toml, lake-manifest.json, lean-toolchain)

Infrastructure:
- Add 4-Infrastructure/shim/: 17 Python probes (RRC manifold, Sidon kernel,
  Wannier, arxiv harvest, math_symbols DB, coverage density, geometric entropy)
- Add 4-Infrastructure/NoDupeLabs/: Node server + package files
- Add 6-Documentation/docs/specs/DP_RRC_RECEIPT_ENCODING_SPEC.md
- Add fix_offloat.py

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-06-18 16:53:23 -05:00
allaun
8cd2fa02f3 feat: NK-Hodge-FAMM formal axiom + Lonely Runner Betti mapping + numerical Betti tracker + vorticity resolution 
Four interconnected solves:

1. NKHodgeFAMM.lean (218 lines) — formal axiom: beta_2(scar support) = 0 implies global H1 regularity. Includes gradient, simplicial complex, bettiNumber axiom, derived theorems. Build: 8598 jobs, 0 errors.

2. lonely_runner_betti_mapping.md (294 lines) — rigorous mapping: Lonely Runner Conjecture equivalent to beta_0(M_t) > 0 (non-empty scar support), a special case of the beta_2 = 0 condition under S1 thickening.

3. betti_tracker.py + test_betti_tracker.py — numerical Betti tracker via gudhi cubical persistence. Computes beta_0, beta_1, beta_2 from velocity field gradient norms. 7 unit tests pass including spherical void detection.

4. cole_hopf_vorticity_resolution.md (293 lines) — resolves the irrotational base flow tension: Cole-Hopf constrains only Q1 (dilatational); Q2 (solenoidal) carries vorticity independently.
 2026-06-16 22:14:10 -05:00
allaun
475f6319ea chore(repo): push local 768-commit branch state onto clean remote baseline 
This squashes all local history (768 commits) onto the scrubbed PR #90
baseline. Individual commits were lost during filter-repo corruption;
the working tree content is preserved intact.

Build: N/A (working tree state only)
 2026-06-15 22:46:50 -05:00
Brandon Schneider
bc9b592dbe docs(infra): update public documentation for virtualized DMA compute backplane 
- Promote the expanded Virtio-Net Packet-as-Computation (PIST) and QEMU graphics backplane spec from the artifacts directory to 6-Documentation/docs/specs/.
- Update root README.md to highlight virtualized DMA computation fabrics.
- Expand 6-Documentation/INFRASTRUCTURE.md to detail host GPU/APU auto-profiling, lossless color ranges, and framebuffer packing shims.
- Keep AGENTS.md aligned with core surfaces.

Build: 3313 jobs, 0 errors (lake build)
 2026-05-30 19:51:24 -05:00
Brandon Schneider
e5fb0a5f4d chore: commit accumulated working tree changes 
Lean: update Semantics modules, add new numerics/physics data files
Hardware: update FPGA bitstreams (tangnano9k_uart_loopback)
Infra: k3s-flake tests, netcup-vps configuration, VCN compute substrate
Docs: ARCHITECTURE, specs, citation updates
 2026-05-30 00:10:02 -05:00
Brandon Schneider
e203a5215e feat(hardware): Emergency Boot Witness reference implementation 
Add compiling Lean modules and Python shim for the Geometry Emergency
Boot Witness specification (graphene supercapacitor geometric encoding).

Lean modules (Semantics.Hardware.EmergencyBoot.*):
- EmergencyBootTypes: HexCoord, CapClass, OpticalPath, VoltageDifferential,
  GrapheneMemristor, HybridOpticalPath, material properties
- EmergencyBootState: PowerState, SolarPowerState, ScanState, seed assembly,
  emergency boot state machine with 6502 calculator efficiency targets
- EmergencyBootShell: Command opcodes, status byte encoding, process
  definitions, executeCommand dispatch

All use Q16_16 fixed-point arithmetic (no Float in compute paths).
Verified theorems: utilizationWithinBounds, powerFailureMonotonic,
commandOpcode_roundTrip.

Python shim (4-Infrastructure/hardware/emergency_boot/):
- EmergencyBootEngine simulating FPGA geometric scan and seed extraction
- Demo CLI showing power failure → self-powered calculator mode →
  geometric scan → seed assembly flow

Build: 3302 jobs, 0 errors (narrow target), 3313 jobs, 0 errors (Compiler)

Generated with [Devin](https://cli.devin.ai/docs)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-05-27 23:00:39 -05:00
Brandon Schneider
6679908f7b fix(arch): enforce Lean-first hierarchy across project — remove all secondary-Lean language 
Project-wide sweep to find and fix every place Lean was treated as secondary,
optional, or subordinate to Python/Rust. The invariant: Lean is the source of
truth. Python and Rust are extraction targets only; they contain no logic, no
invariant checks, no decisions.

Changes:

1-Distributed-Systems/ene/src/lib.rs
  - CRITICAL: Remove 'Rust is the canonical implementation language for
    operational components' — replace with correct extraction-target framing

1-Distributed-Systems/agents/claw/README.md
  - Remove 'canonical implementation lives in rust/' and 'source of truth is
    ultraworkers/claw-code' blanket claims — scope to CLI binary only;
    add Research Stack domain-logic note (Lean is source of truth per AGENTS.md)
  - 'canonical Rust workspace' → 'Rust workspace (I/O extraction target)'

1-Distributed-Systems/agents/claw/src/Tool.py
  - 'Python-first porting summary' → 'Lean-to-Python extraction summary'

1-Distributed-Systems/agents/claw/src/projectOnboardingState.py
  - python_first: bool = True → lean_first: bool = True (Lean always leads)

6-Documentation/docs/specs/ENE_MEMORY_ATLAS_SPEC.md
  - CRITICAL: 'Python first (reference) ... Lean-formal next' → correct order:
    Lean specification first, Python extraction shim, Verilog hardware extraction

6-Documentation/docs/recovered/geocognition_equation_types_map.mmd
  - 'Lean owns logic; Rust owns boundary' → 'Lean owns all logic and decisions;
    Rust is boundary shim only'

6-Documentation/docs/semantics/HYPER_DIMENSIONAL_PHYSICS_INTRO.md
  - 'Python implementation ... Lean formalization' → 'Lean specification (source
    of truth) ... Python extraction shim'

6-Documentation/docs/geometry/GEOMETRY_TAXONOMY_FOR_NLOCAL_ADAPTATION.md
  - 'reference specification for the Python implementation' → 'source of truth;
    Python is an extraction shim against this spec'

6-Documentation/docs/protocols/TM_MCP_SPECIFICATION.md
  - 'Python reference implementation' → 'Python extraction shim' (×2)

5-Applications/scripts/snn/README.md
  - 'deterministic Python reference' → 'Python extraction shim / golden-vector
    harness'; add TODO(lean-port) note; RTL must match 'Python shim (pending
    Lean golden vector)' not 'Python reference'

6-Documentation/docs/METAPROBE_APPROACH.md
  - 'DeltaGCLCompression.lean — Lean implementation / scripts/delta_gcl_encoder.py
    — Python reference implementation' → Lean is source of truth / Python is
    extraction shim

Generated with Devin (https://cli.devin.ai/docs)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-05-26 22:40:03 -05:00
Allaun Silverfox
6d73a1b251 docs(avm): strengthen float prohibition and documentation requirement 2026-05-26 16:02:05 -05:00
Allaun Silverfox
1b6f58c5ce docs(avm): add stripping policy and bad-code elimination rules 2026-05-26 15:59:41 -05:00
Allaun Silverfox
3b4ee97588 docs(avm): redefine AVM as Lean-only ISA with adapter shims/backends 2026-05-26 15:58:25 -05:00
Brandon Schneider
073a70eb86 WIP: accumulated changes 2026-05-25 16:24:21 -05:00
Brandon Schneider
546783faaf docs: add external research import maps 2026-05-20 18:41:21 -05:00
Brandon Schneider
fd863af6fd Expand devcontainer with full Python stack, add MCP servers (Notion/AWS), strengthen Lean theorems 
- .devcontainer/Dockerfile: add PostgreSQL client libs, OpenSSL/libffi headers, gfortran/BLAS for scipy, rclone; install full Python dependency set (boto3, psycopg2-binary, fastapi, uvicorn, notion-client, httpx, pytest, numpy, scipy, etc.) in uv-managed venv; add rclone S3 gateway init script as ENTRYPOINT
- .devcontainer/devcontainer.json: switch from build to pre-built image (localhost/research
 2026-05-19 01:52:14 -05:00
Brandon Schneider
de06a85a83 Quarantine sorry blocks, fix RcloneIntegration proof, add ENE wiki re-ingest and ZFS setup. 
Lean sorry audit (lake build passes, 3539 jobs):
- FixedPointBridge: 10 sorrys quarantined with TODO(lean-port) — all blocked on
  Float→Q bridge lemmas (Q0_16/Q16_16 round-trip error bounds)
- HyperbolicStateSurface: 3 sorrys quarantined — need Q16_16.sqrt error-bound
  and Q16_16.add_pos_of_pos lemmas
- CostEffectiveVerification: 1 sorry quarantined; also fixed pre-existing
  struct/structure typo, Array.Repr, Real.abs syntax, and Bool/Prop mismatch
- MMRFAMMUnification: 1 sorry quarantined — Array.foldl induction lemma missing
- WaveformTeleport: constantWaveformAtFixedPoint_base native_decide was
  numerically false; replaced with sorry + TODO(lean-port)
- RcloneIntegration: startTask_pending_non_increasing PROVED — only sorry fully
  closed, using List.partition_eq_filter_filter + List.filter_sublist
- DiffusionSNRBias, GPUVerificationMetaprobe, QFactor, SSMS: already properly
  quarantined; verified build passes

Infrastructure additions:
- ene_wiki_body_reingest.py: 5-source priority resolver for ene.wiki_revisions
  text="" gap (TiddlyWiki → filesystem → Notion → package description → stub)
- zfs-pool-setup.sh: stackcache pool (500G sparse vdev) with hot/warm/cold
  thermal-zone dataset hierarchy; requires reboot to 7.0.9-1-cachyos kernel

Docs:
- ROADMAP.md: mark Lean→Verilog/FPGA targets as LONG-TERM in Phase 6
- UNIFIED_SIGNAL_ARCHITECTURE.md: add FPGA-column deferral notice

Generated with [Devin](https://cli.devin.ai/docs)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-05-18 23:01:44 -05:00
Brandon Schneider
5a763468c9 integrate infrastructure config, axiom cleanup, and documentation updates 
- cupfox-config.nix: add Open WebUI container with chat.researchstack.info proxy,
  gather-metrics service/timer, rclone, and tmpfiles for persistent storage
- Lean semantics: reduce axiom count from 109 to 18 across 10 files;
  FixedPoint now 0 axioms, 0 sorries with 12 theorems
- Documentation: update AGENTS.md with current axiom/sorry counts and
  FixedPoint status; refine bind signature
- Add topology scripts, CGA/FAMM/GeneticOptimizer/MMRFAMM Lean modules,
  devcontainer config, MEMORY.md, and Modelfile
 2026-05-17 12:03:19 -05:00
Brandon Schneider
a99e839bab Track remaining source and documentation inventory 2026-05-11 22:18:31 -05:00
Brandon Schneider
29f9b78b6d Stage stack solidification source slice 2026-05-11 22:08:10 -05:00
Allaun Silverfox
c6e2d33ef3 docs: add LLM refinement protocol 2026-05-07 08:21:35 -05:00
Brandon Schneider
0cf775c80e collapse: prover orchestration layers, FAMM verilator harness, swarm topological prober, spec sheets, virtual FPGA system tests, merge conflict resolution 
- Prover-Integrated Orchestration Layers (L0-L3): Goedel-Prover-V2 watchdog, BFS-Prover-V2 swarm consensus, bf4prover topology adaptation
- FAMM Verilator benchmark: uniform vs preshaped delay comparison (4.4x speedup)
- Swarm topological device prober: 11 agents probing traces, caps, delays, errors, vias, PDN
- Spec sheet puller: 10 components with key params and topological relevance
- Virtual FPGA system tests: 6/6 passed, 134K ops/s throughput
- Fixed merge conflicts in AI-Newton test_experiment.ipynb
 2026-05-06 23:42:01 -05:00
Brandon Schneider
0709b298b3 Consolidate research stack updates 2026-05-05 21:09:48 -05:00
Brandon Schneider
5f88abf618 initial: sovereign research stack (consolidated, weightless, and lfs-optimized) 2026-05-04 18:11:36 -05:00