Commit graph

4 commits

Author SHA1 Message Date
00e9eed399 fix(lean): complete projectionOrdering proof in GeometricCompressionWorkspace
Replace the TODO(lean-port) sorry with a complete proof of the
projectionOrdering theorem: for positive SourceValue pairs s1 < s2
with s2 ≤ maxExpected, projectToCoding preserves strict ordering
of the Q0_64 values.

The proof uses Nat-only arithmetic (no Float) and handles two cases:
  - a2 < d: both values fit in Q0_64 range, ordering follows from
    monotonicity of integer division
  - a2 = d: a2*s/d = s clamped to q0_64MaxRaw; a1*s/d < q0_64MaxRaw
    via the key inequality (d-1)*s < (s-1)*d

Build: 8598 jobs, 0 errors (lake build)
2026-06-18 15:06:50 -05:00
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
4a72ad1183 feat(lean): port pist_trace_classify_mcp spectral logic to Semantics.PIST.Spectral
## New module: Semantics.PIST.Spectral

Ports the two domain-logic functions from pist_trace_classify_mcp.py
that were previously executing in unverified Python:

### classify_tactic_from_name → classifyTacticFromName
- `TacticFamily` inductive (rewrite, normalization, arithmetic, induction,
  algebraic, case_analysis, discharge, reflexivity, unknown)
- Pure string-lookup; 5 executable witnesses confirm all branches.

### compute_spectral → computeSpectral
- `isqrt` — integer Newton's method for floor(√n); 4 witnesses.
- `powerIteration` — Q16_16 fixed-point dominant eigenvalue via power
  iteration with Rayleigh quotient; identity-matrix witness = 65536.
- `SpectralProfile` structure — 10 fields (matrix_size, rank,
  spectral_gap, density, trace_val, frobenius_norm, laplacian_zero_count,
  adjacency_eigenvalue_max, laplacian_eigenvalue_max, singular_value_max).
- `computeSpectral` — symmetrize → lap → powerIteration → shift-deflation
  for second eigenvalue → AᵀA for singular value; 3 witnesses on 2×2 fixture.

No Float in any compute path. All magic constants documented with formulas.

## Other changes
- Semantics.lean: add `import Semantics.PIST.Spectral`
- AgenticOrchestration.lean:163: expand bare `-- TODO(lean-port):` label
- pist_trace_classify_mcp.py: update PARTIAL BOUNDARY comment to name
  the Lean module that now owns spectral logic

## Build baseline
  lake build Compiler → 3311 jobs, 0 errors

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

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-05-26 23:25:50 -05:00
Brandon Schneider
5f88abf618 initial: sovereign research stack (consolidated, weightless, and lfs-optimized) 2026-05-04 18:11:36 -05:00