bgilbert1984/SCYTHE: SCYTHE @ spectrcyde.com https://github.com/bgilbert1984/SCYTHE
Cite the SCYTHE catalogue work as a chain of primary repository records, because the design deliberately evolved through proposals, implementation, adversarial certification, live evidence, and authorization decisions.
- Receiver-spur identification method — §5.21. SCYTHE Project, “docs(5.21): propose receiver-spur identification — and say what it cannot buy,” GitHub PR #77, merged as d5bbbba1 (2026). This is the foundational citation for the retune-slope model, the distinction between received emissions and internal mixing products, the requirement for at least three retunes, termination as a discriminator, and the rule that stability classifies rather than excludes.
- Experimental design and frozen catalogue parameters — §5.22. SCYTHE Project, “docs(5.22): propose the capture plan — estimand first, and two things that exposes,” GitHub PR #78, merged as 78642344 (2026). This is the primary citation for K=64, R=8, retune deltas {50,100,200} kHz, slope tolerance 0.01, ≥7/8 persistence, ≥10 dB margin, counterbalanced/repeated visits, temporal blocking, and the conservative reference-comb cap. It explicitly derives the ±100 ppm harmonic cap of 3.
- Catalogue slope analysis implementation. SCYTHE Project, “§5.21 catalogue analysis: the slope is measured, the tolerance governs it, and the catalogue answers entry 10,” GitHub PR #118, merged as 4091d22e (2026). This implemented SpurSlopeEstimate, stored the measured slope/intercept/residuals, enforced classification support, and made the catalogue govern where THERMAL_NO_INPUT and RECEIVER_SPURS may be captured.
- Adversarial certification of the slope/catalogue analysis. SCYTHE Project, “§5.21 S-series: 21 controls over the catalogue analysis, certified at 94e7840b,” GitHub PR #120, merged as 007f8e5d (2026). The second formal sweep certified 21/21 controls, including slope tolerance, declared retunes, folding, residuals, classification support, thermal-spur exclusion, and reconstruction behavior.
- Reference-comb enforcement. SCYTHE Project, “§5.21 entry 17: the reference-comb match governs the reference class; live catalogue protocol for review,” GitHub PR #121, merged as 39da055e (2026). This added ReferenceComb, harmonic declaration, derived RF position, the n·f_ref·ppm match window, the harmonic cap, and refusal of unsupported reference-class claims. It also records the important consequence that ±100 ppm limits the cap to 3, eliminating reference-class claims in UHF.
- Adversarial certification of the reference-comb checks. SCYTHE Project, “§5.21 S-series: S22..S35 over the reference-comb match, certified at 21de3d9c,” GitHub PR #122 (2026). The certified run exercised 33 total §5.21 controls, including harmonic presence, cap application, comb matching, RF-position derivation, catalogue-wide matching, and disagreement/refusal paths. The PR is currently represented as open in GitHub metadata, so I would cite it specifically as the certification record, not as a merged-main milestone.
- Live catalogue machinery. SCYTHE Project, “§5.21 catalogue runner: retention estimator, live sequencer, authorisation record, protocol governance,” GitHub PR #124, merged as 5a425aca (2026). This is the implementation citation for the Hann-windowed persistence estimator, 7-of-8 retention, byte-exact live tuner sequencer, catalogue writer, and post-capture auditor. It also preserved the crucial refusal boundary: no catalogue means no manufactured CLEAN verdict.
- Pre-catalogue empirical evidence. SCYTHE Project, “Three Windows, One Tone: 100.8 MHz, Persistent Across Reconnect and Power Cycle,” GitHub PR #125, merged as 7e38abf0 (2026). This is the experimental citation for the three 120-second terminated-input captures, SHA-256-preserved transfer, the ~100.7999688 MHz feature, identical FFT bin at 7.8125 Hz resolution, and ≥0.998 spectral-shape cross-correlation across initial, reconnect, and power-cycle observations. It explicitly keeps those observations as PRE_CATALOGUE_EVIDENCE, not catalogue entries.
- Formal authorization/no-go boundary. SCYTHE Project, “§5.21 catalogue authorization decision,” GitHub PR #126, merged as 8361e26b (2026). This froze the distinction between empirical observations and catalogue authority, explicitly kept reference_hz and reference_ppm undeclared, prohibited result-selected references, and held the K=64/R=8 run at NO-GO pending unit-specific evidence.
- Acquisition reliability / false-green protection. SCYTHE Project, “ops: codify ss-guard rtl_tcp watchdog,” GitHub PR #127, merged as 9a5a1713 (2026). This is the operational citation for the read-only RTL0 + actual-I/Q-flow health probe, /proc/net/tcp inspection, 3/hour restart limit, zero-byte-command assertion, and rule that watchdog recovery invalidates rather than papers over an interrupted catalogue visit.
- Final reference authority and catalogue GO. SCYTHE Project, “§5.21: reference declaration (28.8 MHz, 100 ppm uncalibrated) + catalogue GO,” GitHub PR #128, merged as d68e034e (2026). This is now the canonical citation for UNIT_INSPECTED_MODEL_CORROBORATED, the physical Y6 inspection/photos and hashes, reference_hz = 28,800,000, the conservative reference_ppm = 100, explicit exclusion of the 100.8 MHz observation from the authority chain, harmonic cap 3, and authorization of the K=64/R=8 catalogue schedule.
- For a short bibliography, cite just #77, #78, #118, #121, #124, #125, and #128. Those tell the complete scientific story: method → experimental design → mathematical implementation → reference model → acquisition machinery → empirical shakedown → evidence-backed authorization. The S-series and watchdog citations then serve as supporting verification/operations references.
Ben Gilbert 409-334-4829