Trisha validation records

The coordinated CPU release is published: Trident 0.3.0, Trisha 0.3.0 and Joy 0.5.0. Current release report records all six native targets, exact sources/binaries and completed acceptance gates.

Historical validation records

These dated reports preserve earlier candidates and findings; their open-item status is superseded by the published release report.

• Native release execution: current native builds, Windows fixes, installed proofs and exact CI/source identities.

• Release platform review: Windows requirement, proposed six-target matrix, native validation coverage and portability gaps.

• Compiler scratch/RAM blocker: two confirmed source memory corruptions, preservation design and repair status. FINAL6 is blocked.

• RAM preservation repair: stack cleanup, deep access and assembly repairs; complete execution baselines, eight fresh CLI proofs and exact before/after receipts. Fresh frozen-platform proof gates remain open.

• FINAL6 coordinated candidate: next exact source snapshot after ownership, file-reader, formal and dependency changes; native build and proof gates are tracked individually.

• FINAL5 coordinated candidate: current frozen native builds, installed proofs, reproducible archives and bidirectional proof verification; complete baseline gate status is recorded separately.

• FINAL5 Linux evidence: native installed/node admission and freshly unpacked cross-platform verification receipts.

• Complete baseline gate: exact candidate/input inventories, actual verified proof events and the corrected Python startup.

• Current release tooling:24 actual packaging, source-inventory and proof-event gate tests.

• Experimental tagged CCS proofs: two real proofs with topology/input/cost binding; no production protocol integration claim.

• Failed FINAL4 candidate: preserved static loop-index failure, diagnosis and the correction included in FINAL5.

• Workspace and CLI validation: exact test totals, repaired PLUMB checks and separately invoked expensive proof gate.

• Platform checks: Rust1.89 CPU and Linux cross-checks, with compilation distinguished from execution.

• Recursive proofs: fresh Triton 7 source/hand STARKs, complete inventory accounting and actual prover memory observations.

• Linux runtime validation: isolated native toolchain, source archive builds and installed runtime gates.

• Source rehearsal: deterministic archive, exact provenance and native candidate binary hashes.

• Source verification: full source/vendor inventories, symlink boundaries and failure cleanup around candidate builds.

• Installed smoke: complete macOS proof and rejection checks bound to installed binary and public-fixture hashes.

• Binary packaging: deterministic archives restricted to the exact installed binaries and fixtures that passed the full smoke.

• Local Neptune node: genuine SingleProof, actual authenticated CLI submission, exact mempool admission and cleanup.

• Neptune adapter review: independent intent, output, canonical JSON and authenticated transport inspection.

• Native proof inputs: canonical wire/claim/input validation, resource limits, malformed transcripts and remediation.

• Neptune CLI: selected network and wallet/node process behavior with isolated fixtures.

• Mining compatibility: pinned consensus oracle, race/budget/template findings and subsequent repair receipts.

• GPU claims and older roadmap: unverified historical performance/status claims, retained as history.

Baseline fixtures and their oracle provenance are described in baselines/triton. Execution and full source/hand proof coverage are separate gates. New evidence must preserve the43-file inventory and disclose any implementation-specific public statements.