LINDY Weekly Program Receipts — Deterministic Trajectory Identity and Cage Containment, 2026-10-15 to 2026-10-31

Weekly program receipts for two deterministic checks, run on a Friday 22:30 CT cadence against frozen preregistrations, weeks 2026-10-15 through 2026-10-31. LINDY-B — trajectory identity and canary scan. Recomputes the trajectory-to-canonical hash and compares it to the pinned baseline ca575bb9b781c47908bb13c24ba5739021e1fc15f33bc245257836d866f93d50, verifies lineage 106/106 and 2646 crossings, and reconstructs a closure report against a 1e-04 budget. LINDY-C — cage containment falsifiers FC1-FC5. Instrument integrity over 12 frozen hashes, containment truth, a positive control admitted 8/2/2, 8/8 adversarial candidates rejected with zero side effects, and journal invariant 8 == admitted with canonical journal hash d51cd63369c913eb7090a09040f73104b6b379d442156c1b13e6adf1f5d86b98. Result. Every PASS in the window reproduces the pinned baseline byte-for-byte: zero drift in the LINDY-B trajectory hash and no change in the LINDY-C canonical journal hash. LINDY-C carries five clean passes across the window against two disclosed path-defect misses. Disclosures. (1) Two path-defect misses occurred (a displaced telemetry pack in LINDY-B, an absent input pack in LINDY-C). Neither was a result about the system under test; each aborted the run before candidates executed, was disclosed at the time, and was repaired the same night without modifying harness or corpus bytes. (2) The 2026-10-23 LINDY-B row is absent from LINDY_B_RUNLOG.md; the pass is recorded in the project ledger entry 0324 with its slot signature and hash evidence. The omission is a recording defect and is left visible rather than back-filled. (3) Synthetic corpus only: no production, customer, or personal data, and no PHI, appears in this bundle. No operator chat text is included. (4) Prototype research evidence, not a production-hardened product; no production warranty is implied. Dual Mesh Benchmark: NOT MEASURED for these runs. LINDY-B and LINDY-C are deterministic trajectory and containment checks rather than LLM capability benchmarks. Measured DMB columns are published separately (e.g. DOI 10.5281/zenodo.23609232 frontier leg; DOI 10.5281/zenodo.23345241 DeepSeek leg, scoped measurement). No on-chain anchoring is part of this record; receipts are git-committed and hash-addressed. Bundle sha256 77acc1459be745c1b1a466460a8f89520e3819b2b036481b861d7b7bf35c4d4a.