Bounded endogenous-form codec. Tunnels, spirals, lattices, symmetry metadata, and profile fields round-trip exactly within the declared release corpus.

ZPE-Mental is one of 17 independent encoding products in the Zer0pa portfolio, each scoped to its own domain; Mental covers the endogenous visual modality and makes no claim beyond it.

The product is useful now and improving continuously. The current surface is prepared for exact bounded-form replay on the corpus committed here and for fresh falsification from a clean outsider shell. It does not claim broad cognition decoding, language understanding, autobiographical memory recovery, or general decoding of non-visual mental content.

Source: proofs/artifacts/mental_release_matrix.json; validation/results/pytest.xml; proofs/manifests/CURRENT_FALSIFICATION_PACKET.md

Broad cognition decoding or any general decoding of non-visual mental content.; Recovery of autobiographical memory, legal or moral reasoning, theorem proving, or language planning.; Clinical diagnosis, therapeutic interpretation, or any medical use.

• Exact round-trip preservation for the bounded endogenous visual form corpus in validation/corpora/endogenous_forms.json (4 forms: tunnel, spiral, lattice, cobweb).
• Parity between the Rust-native encoder/decoder and the pure-Python reference implementation on the same bounded corpus (BASELINE_DELTA = 0.00; byte-level word equality asserted in CI).
• Stable profile and symmetry handling for both COMPASS_8 (8 directions at 45° increments) and D6_12 (12 directions at 30° increments, enabling exact D6 rotational symmetry) direction surfaces included in the release corpus.
• Deterministic collapse of non-visual prompts to a bounded fallback form instead of broad-cognition retention (NON_VISUAL_SEMANTIC_RETENTION = 0.00; NON_VISUAL_ALIAS = 1.00).
• Compact 20-bit wire format supporting raw and RLE direction encodings, per-stroke frame-index and delta-time metadata, all within the same word space. RLE is the default transport (Augmentation Phase 2). Proof: zpe_mental/pack.py, src/lib.rs.

• Broad cognition decoding or any general decoding of non-visual mental content.
• Recovery of autobiographical memory, legal or moral reasoning, theorem proving, or language planning.
• Clinical diagnosis, therapeutic interpretation, or any medical use.
• A broader shared stack or any decoding coverage outside endogenous visual forms.

Rows retained from the previous expanded ## Key Metrics table. The public product page uses the first four rows only.

ZPE-Mental encodes endogenous visual forms (tunnels, spirals, lattices, cobwebs per the Klüver form taxonomy). This modality has no incumbent codec market and no established external encoder/decoder. The closest domain — fractal geometry encoders or L-system compressors — generate fractals from rules rather than encode/decode closed perceptual forms; the comparison would be apples-to-oranges. The lane's product claim is exact roundtrip fidelity on the bounded form corpus (FORM_EXACT = 1.00), not byte compression.

python3 - <<'PY'
import sys
raise SystemExit(0 if sys.version_info >= (3, 11) else "Python 3.11+ is required")
PY
cargo --version
python3 -m venv .venv
. .venv/bin/activate
python -m pip install --upgrade pip
python -m pip install ".[test]"
python scripts/validate_release.py --output proofs/artifacts/mental_release_matrix.json
pytest tests -q --junitxml validation/results/pytest.xml

This section captures the active lane priorities — what the next agent or contributor picks up, and what investors should expect. Cadence is continuous, not milestoned.

• Klüver-form codec — market and admission path — Zero-Base Scientific Thinking — GPD Research and Planning Pending. Lane has no incumbent codec market for endogenous visual forms. Active research into market definition (perceptual encoding, neuroscience tooling, altered-states documentation, art-tech) and the certified-subset admission path before corpus expansion past the four bounded forms.