壁相对局部晕超压 ~3×（L13：WSW 2.95 ± 0.16 → C2 后 2.87；ENE ~2.6 → 2.83），因此它们不可能是被致密晕静态约束的壳。P2（sol）在固定 nH(47) = 5.5e-5 下检验了静力学假设：0.2 keV 接近 HSE 形状（M200 = 2.93e12 M⊙，T_vir200 = 0.132 keV，M(<50 kpc) = 1.7e9 M⊙），但 16 个 HSE 晕模型给出 WSW 平均 q = 2.16–3.78（从不接近 1）而斜率 −0.45 到 +0.60 —— 静力学"持平"不是普适预言。壁的冷却时间很长（APEC t_cool 5.9–14.0 Gyr，t_cool/t_ff 27–49，WSW FXTwide 实测 bin），所以壁是缓慢弛豫中的准静态储层。决定性的张力在亮度而非重子总量：平衡晕（n47 = 1.4e-4）预测 S(8.5′)−S(22.5′) ≈ 3.95 PW（β,0.2 模型），与 FXT 实测 −0.1 ± 1.4 PW 不容（E084；P2 复核）。L16（条件性）：若壁光滑且热支撑，47 kpc 总晕压 ≲ 3 P_X。
The walls are over-pressured by ~3× against the local halo (L13: WSW 2.95 ± 0.16 → 2.87 after C2; ENE ~2.6 → 2.83), so they cannot be a shell statically confined by a dense halo. P2 (sol) tested hydrostatics at fixed nH(47) = 5.5e-5: 0.2 keV is close to the HSE shape (M200 = 2.93e12 M⊙, T_vir200 = 0.132 keV, M(<50 kpc) = 1.7e9 M⊙), but 16 HSE halo models give a WSW mean q = 2.16–3.78 (never near 1) with slopes from −0.45 to +0.60 — a hydrostatic "flat" profile is not a generic prediction. The wall cooling times are long (APEC t_cool 5.9–14.0 Gyr, t_cool/t_ff 27–49, WSW FXTwide measured bins), so the walls are a quasi-static reservoir relaxing slowly. The decisive tension is brightness, not total baryon mass: the balance halo (n47 = 1.4e-4) predicts S(8.5′)−S(22.5′) ≈ 3.95 PW (β,0.2 model) against the measured FXT value of −0.1 ± 1.4 PW (E084; P2 recheck). L16 (conditional): if the walls are smooth and thermally supported, the total halo pressure at 47 kpc is ≲ 3 P_X.
来源Source: P2 REPORT.sol.md; E084_s09_halo_limit_robustness.md; E098a (L13, L16); REPORT.hermes.md Table 3
INFERRED（薄壳模型）若把壁的热压当作激波后压强，Rankine–Hugoniot 关系给出 Mach ≈ 1.5–1.6（WSW 1.60、ENE 1.48；激波速度 340–370 km/s，接触面/壁速度 140–170 km/s，压缩 1.7–1.85），预测的外鞘 kT 0.29–0.32 keV 未被探测也未排除（L18）。Li (2026) 维度化标度给出 E0 2.2e57（WSW）/ 4.2e57（ENE）erg 每瓣、t_P 72/114 Myr —— 泡在 t_P 之前、Stage II–III 过渡附近。薄壳瓣（45° 锥、分层晕、含引力）匹配 80–100 kpc 处 Mach 1.60 的历史表：100 Myr 情景 E0 = 3.40–4.94e57 erg 每瓣、注入功率 L_on = 1.08–1.57e42 erg/s 每瓣（两瓣为 SN 功率 8e41 的 2.7–3.9×），结束于 51–76 Myr 前；7–9e41 erg/s 是对整个泡龄 151–176 Myr 的平均功率（A1 审计纠正了 E098 行 61 与 e098a_figure.py 98–100 的措辞；1,240 项算术检查通过）。连续注入到今天被现今风上限排除（×50）。R1 红队（致命点）：Mach 是"假设壁压 = 激波后压强"推定的，不是测量的速度或跳变；对唯一激波历史的断言是致命的，对条件性模型是重大的。能量下限 ≥ 3.4e57 只在 Mach 1.60 + 引力分支成立；全部能量 ∝ P_halo(47 kpc)/4.1e-14。
INFERRED (thin-shell model) If the wall thermal pressure is identified with the post-shock pressure, the Rankine–Hugoniot relation gives Mach ≈ 1.5–1.6 (WSW 1.60, ENE 1.48; shock speed 340–370 km/s, contact/wall speed 140–170 km/s, compression 1.7–1.85); the predicted 0.29–0.32 keV outer sheath is neither detected nor excluded (L18). The dimensional Li (2026) scalings give E0 2.2e57 (WSW) / 4.2e57 (ENE) erg per lobe and t_P 72/114 Myr — the lobes are before t_P, near the Stage II–III transition. A thin-shell lobe (45° cone, stratified halo, with gravity) matched to Mach 1.60 at 80–100 kpc gives, for the 100 Myr episode, E0 = 3.40–4.94e57 erg per lobe and injection power L_on = 1.08–1.57e42 erg/s per lobe (2.7–3.9× the 8e41 erg/s SN power for two lobes), ending 51–76 Myr ago; 7–9e41 erg/s is the power averaged over the full 151–176 Myr bubble age (the A1 audit corrects the wording of E098 line 61 and e098a_figure.py 98–100; 1,240 arithmetic checks pass). Continuous injection to today is excluded by the present-wind limit (×50). R1 red team (fatal point): Mach is inferred by assuming wall pressure = post-shock pressure, not measured from a velocity or a jump; this is fatal to a unique shock history and major to the conditional model. The energy floor ≥ 3.4e57 holds only on the Mach 1.60 + gravity branch; all energies scale as P_halo(47 kpc)/4.1e-14.
来源Source: E098 record + A1 REPORT.sol.md (audit); R1 REPORT.sol.md rank 5; R1 REPORT.astra.md (injection vs mean power)
CONDITIONAL 以两瓣 6.8e57 erg（100 Myr 恒定功率）为靶：现今 SFR 2.4–8.3 M⊙/yr 的 SN 功率为 8.3e41–2.9e42 erg/s，需要耦合 0.75–2.6（SFR 5.2 时 1.2–1.3 —— 星暴在 100% 耦合下也勉强够）；射电喷流 4–6e41 erg/s 需要 3.6–5.4（单独不可能）；AGN 辐射在 MIR L_bol 4.1e43（Fernandez et al. 2023）下需 5% 耦合，在 CT X 射线 L_bol 3.4e42（Brightman et al. 2016）下需 64%。文献中没有 50–300 Myr 的核/盘恒星形成史（NOT TESTABLE）；已发表的年龄是风/泡/射电电子年龄（Myr 尺度）。P1（氧预算）独立指出：能量路线在 ε = 0.3/0.1 时需要 20–23/62–68 M⊙/yr 的恒星形成（两瓣），是现今 SFR 的 2.4–28× —— 能量收支与现今 SFR 在考虑氧之前就已冲突，需要更高的历史 SFR 或 AGN。结论：AGN 不是必需的，但能量来源也未确定；两个源都允许、都未证实。
CONDITIONAL With a two-lobe target of 6.8e57 erg over 100 Myr at constant power: the SN power of the present SFR 2.4–8.3 M⊙/yr is 8.3e41–2.9e42 erg/s and needs coupling 0.75–2.6 (1.2–1.3 at SFR 5.2 — the starburst barely supplies it even at 100% coupling); the radio jet at 4–6e41 erg/s needs 3.6–5.4 (cannot do it alone); AGN radiation needs 5% coupling at the MIR L_bol 4.1e43 (Fernandez et al. 2023) or 64% at the CT X-ray L_bol 3.4e42 (Brightman et al. 2016). No 50–300 Myr nuclear/disk star-formation history exists in the literature (NOT TESTABLE); published ages are wind/bubble/radio-electron ages on Myr scales. P1 (oxygen budget) independently notes: the energy route at ε = 0.3/0.1 needs 20–23/62–68 M⊙/yr of star formation for two lobes, 2.4–28× the present SFR — the energy budget conflicts with the present SFR before oxygen is even considered, requiring a higher past SFR or an AGN. Conclusion: an AGN is not required, but the source is not identified; both sources are allowed and neither is demonstrated.
来源Source: L1 REPORT.luna.md (energy table); P1 REPORT.sonnet.md (T3b, T4); session 2026-10-15.md (arithmetic checked)
INFERRED P1（氧预算）用六组核心坍缩超新星（CCSN）产额（Kroupa IMF）：CCSN 喷出物 O/Fe = 1.4–3.6（太阳单位）；加 SN Ia（Maoz & Graur 2017 DTD，150 Myr）后降到 1.1–2.9。观测的 1T O/Fe（4.9–10.2）没有任何产额组能达到；与太阳比气体的混合只会更低（预测 1.0–2.8）。要达到观测值需要 Fe 亏损 ×1.9–4.8（O/Fe 4.9）到 ×3.6–9.6（10.2），或强 1T 偏差。氧的质量不是问题：WSW 壁（f=1，M ≈ 1e9 M⊙）在 Z_O = 1 时含 M_O ≈ 5.8e6 M⊙，与 3.4e6 个 SN 的氧产量同量级（比值 1.0–3.3，ε = 1，Z_O = 1；ε = 0.3 时 0.31–0.98）。Liu et al. 2026 的 RGS 核心值 0.47 对 EPIC 1T 3.5–5.1（同一核心区，因子 7–11）是 1T 偏差最直接的证据；该大小的偏差会把测量值带回喷出物/混合区间 1–3。绝对 O 丰度不受约束（拟合只约束 O/Fe 与 O 线发射度量；a_O ~ 0.3–0.4 只是提示）。论文 II 必须做的：修复合并似然计数；1T/2T/DEM/1T+CX 模型空间与自由前景（C6）；O VII/VIII 对 Fe-L；固定 kT 下的 O/Fe；XMM/FXT 分裂（2.2–2.6×）的理解；Lodders→Asplund 换算（数值上 ×1.00，P1 已核）。
INFERRED P1 (oxygen budget), with six core-collapse supernova (CCSN) yield sets (Kroupa IMF): CCSN ejecta O/Fe = 1.4–3.6 (solar units); adding SN Ia (Maoz & Graur 2017 DTD, 150 Myr) lowers it to 1.1–2.9. No yield set reaches the observed 1T O/Fe (4.9–10.2); mixing with solar-ratio gas only lowers it further (predicted 1.0–2.8). Reaching the observed values needs an Fe deficit of ×1.9–4.8 (O/Fe 4.9) to ×3.6–9.6 (10.2), or a strong 1T bias. The oxygen mass is not the problem: the WSW walls (f = 1, M ≈ 1e9 M⊙) hold M_O ≈ 5.8e6 M⊙ at Z_O = 1, comparable to the oxygen made by 3.4e6 SNe (ratio 1.0–3.3 at ε = 1, Z_O = 1; 0.31–0.98 at ε = 0.3). The Liu et al. 2026 RGS nucleus value of 0.47 against EPIC 1T 3.5–5.1 (same nucleus, a factor 7–11) is the most direct evidence of the 1T bias; a bias of that size would bring the measured values back to the ejecta/mixed range 1–3. The absolute O abundance is not constrained (the fits constrain O/Fe and the O-line emission measure; a_O ~ 0.3–0.4 is only a hint). What Paper II must do: fix the pooled-likelihood counting; the 1T/2T/DEM/1T+CX model space and a free foreground (C6); O VII/VIII against Fe-L; O/Fe at fixed kT; understand the XMM/FXT split (2.2–2.6×); the Lodders→Asplund conversion (numerically ×1.00, checked by P1).
来源Source: P1 REPORT.sonnet.md (T2, T5, reading); Liu et al. 2026 arXiv:2609.37646; I021 C6; session 2026-10-15.md
MEASURED（E097）投影上轴柱在 < 28 kpc 处不比壁脊暗（WSW 约亮 2σ，因为视线穿过前后壁）；按三维发射率（3-d emissivity）WSW 壁在所有高度都比内部亮 2–3×（η 0.3–0.5 below 28 kpc → ~0 above 39 kpc），但变化只有 1–1.5σ，没有尖锐过渡。内部可以容纳不可见的热薄气体：壁压处 kT ≥ 0.5 keV 的内部给出软带 η ≤ 0.24、硬带低于噪声 —— 这是自洽图像（允许、未探测）；否则壁会在 70–100 Myr 的声波越壁时间内向内扩散。E093：现今风的冲压压力在 ENE ≤ 2–7 kpc、WSW ≤ 3–25 kpc 就低于晕压，推不到 40–80 kpc；无障碍物（HI/Hα/射电在 40–70 kpc 的两轴上都没有已知气体）能制造 M82 式的帽；软强击波帽在测得晕密度下边缘排除，热帽允许。INFERRED（L2 luna）湍流混合层（TML）不能制造 HK2020 的 FUV 丝：以 Tan et al. 2021 的外 ENE 壁条件（P 1.07e-13 dyn cm⁻²、v_shear 150 km/s、T_cold 1e4 K），GALEX 波段线亮度预测为 5.5e-21 erg/s/cm²/arcsec²（每单位面积因子 f_A），即预测/观测 = 2.3e-4 f_A（用 HK2020 §3.1 prose 值 20.4 mag/arcmin²）或 2.3e-3 f_A（用其 Table 2 的 19.8 mag over 16.8 arcmin² = 22.86 mag/arcmin²，即经过 PSF 翼与尘埃改正后的正确比较值）。把整个光滑 38–66 kpc ENE 锥台都算给两个东箱（f_A ≈ 13）也只有 ~0.003/0.03。只靠面积解释残差会过度处理壁质量基准（~30 M⊙/yr、4.7e9 M⊙ in 156 Myr，是参考壁质量的 7–9×）。更正：luna 报告称 HK2020 prose 值与 Table 2 "冲突"是错的 —— prose 值是未减背景的箱平均，Table 2 才是改正后的值（session 2026-10-15 已核，arXiv:2009.09047 §3.1 与 Table 2 注）。
MEASURED (E097) In projection the axis column is not fainter than the wall ridge below 28 kpc (WSW about 2σ brighter, because the line of sight crosses the front and back walls); per 3-d emissivity the WSW walls are 2–3× brighter than the interior at all heights (η 0.3–0.5 below 28 kpc → ~0 above 39 kpc), but the change is only 1–1.5σ with no sharp transition. The interior can hold invisible hot thin gas: an interior at the wall pressure with kT ≥ 0.5 keV gives soft-band η ≤ 0.24 and stays below the hard-map noise — the self-consistent picture (allowed, not detected); without it the walls would spread inward on the 70–100 Myr sound-crossing time. E093: the present wind's ram pressure falls below the halo pressure at ENE ≤ 2–7 kpc and WSW ≤ 3–25 kpc, so it cannot push to 40–80 kpc; no obstacle (no known HI/Hα/radio gas across either axis at 40–70 kpc) can make an M82-like cap; a soft strong-shock cap is marginally excluded at the measured halo density and a hot cap is allowed. INFERRED (L2 luna) Turbulent mixing layers (TML) cannot make the HK2020 FUV filaments: with the Tan et al. 2021 conditions on the outer ENE wall (P 1.07e-13 dyn cm⁻², v_shear 150 km/s, T_cold 1e4 K), the predicted GALEX-band line brightness is 5.5e-21 erg/s/cm²/arcsec² per unit area factor f_A, i.e. predicted/observed = 2.3e-4 f_A (with HK2020's §3.1 prose value, 20.4 mag/arcmin²) or 2.3e-3 f_A (with their Table 2 value, 19.8 mag over 16.8 arcmin² = 22.86 mag/arcmin², the correct comparator after PSF-wing and dust correction). Assigning the whole smooth 38–66 kpc ENE frustum to the two east boxes (f_A ≈ 13) still gives only ~0.003/0.03. Explaining the residual by area alone would overprocess the wall-mass benchmark (~30 M⊙/yr and 4.7e9 M⊙ in 156 Myr, 7–9× the reference wall mass). Correction: the luna report's claim that HK2020's prose value "conflicts" with Table 2 is wrong — the prose value is the box average without background subtraction, and Table 2 is the corrected value (checked in session 2026-10-15, arXiv:2009.09047 §3.1 and the Table 2 note).
来源Source: XMMLineMap E097, E093; L2 REPORT.luna.md; session 2026-10-15.md (HK2020 comparator correction)
CONDITIONAL（P4 sol 报告；Hermes 分支读过 sol 的文件，不是独立检验）NGC 3073（矮 S0，投影 ~55 kpc，PA 249，距 WSW 轴 6–7°；HI 尾 ~28 kpc 背向 NGC 3079）：在两套恒星模型下盘缘恢复压力为 1.69–5.17e-12 dyn cm⁻²（球锚定 9.85e-12），而任何受试的 100–600 km/s 内流（软/热）都达不到剥离判据 —— nH v² 需 ≥ 221 cm⁻³(km/s)²（fiducial GG），最大软内流仅 27.6。现今风的冲压 ≤ 1.8e-14，低于历史弥散尾阈值 5.65e-14 的 1/3。HK2020 假设 n = 5e-4 cm⁻³，比我们的值高约 10×；其"likely, but barely, sufficient"不能对该阈值数值复现。距离文献冲突（17.8–33.7 Mpc），关联与局部遭遇是假设。只有快速内流能扰动其弱束缚的尾气；不能建立当前盘剥离。MEASURED（P5）MCG 9-17-9：物理伴星系（Δv 173–192 km/s，33 kpc HI 桥，Shafi+2015），位于 WSW 壁楔内，点源已扣除；对 L7 与烟囱边缘无影响，但 WWLM 谱区条件性暴露（§3.12）。R/V = 2.1e8 yr 与 E098a 爆发起点 1.5–1.75e8 yr 的重叠只是量级巧合；与 ~100 Myr 风期的因果联系 NOT TESTABLE（投影因子 ~2）。
CONDITIONAL (P4 sol report; the Hermes arm read the sol arm's files, so it is not an independent check) NGC 3073 (a dwarf S0, projected ~55 kpc, PA 249, 6–7° from the WSW axis; an HI tail ~28 kpc pointing away from NGC 3079): the disk-edge restoring pressure is 1.69–5.17e-12 dyn cm⁻² for two stellar models (9.85e-12 for spherical anchoring), and none of the tested 100–600 km/s interior cases (soft or hot) reaches the stripping criterion — nH v² must be ≥ 221 cm⁻³(km/s)² (fiducial Gunn–Gott) while the largest soft-interior product is only 27.6. The present wind's ram pressure, ≤ 1.8e-14, is below even the historical diffuse-tail threshold of 5.65e-14 (about 1/3 of it). HK2020 assumed n = 5e-4 cm⁻³, about 10× our value; their "likely, but barely, sufficient" is not numerically reproducible against that threshold. Literature distances disagree (17.8–33.7 Mpc); the association and a local encounter are assumptions. Only a fast interior can disturb its weakly bound tail gas; ongoing disk stripping is not established. MEASURED (P5) MCG 9-17-9: a physical companion (Δv 173–192 km/s, a 33 kpc HI bridge, Shafi+2015) inside the WSW wall wedge, its point source subtracted; no effect on L7 or the chimney edge, but the WWLM spectral region is conditionally exposed (§3.12). The overlap of R/V = 2.1e8 yr with the E098a episode onset 1.5–1.75e8 yr ago is an order-of-magnitude coincidence only; a causal link to the ~100 Myr wind episode is NOT TESTABLE (projection factor ~2).
来源Source: P4 REPORT.sol.md; P5 REPORT.sonnet.md; S09 FACTS (objects); session 2026-10-15.md (P4 Hermes arm not independent)
R1（sol 主报告 + astra 部分报告，两臂独立）按严重度排序了八个最弱环节，并给出了逐条措辞降级。sol 排名：1 L7 端点（合并 bin，非端点；PN 单独弱）；2 L12 晕（亮度梯度测得，绝对密度/压力依赖球形、热谱、丰度与常数拟合）；3 L11 O/Fe（拟合比值不是绝对富集；pooled 太阳在 2σ 允许）；4 L13 2.95 ± 0.16（误差仅壁归一化；斜率模型依赖）；5 L15 Mach（推定而非测量 —— 对唯一激波史致命）；6 L14–L17（薄壳能量缺壳热；注入功率与寿命平均功率混淆）；7 L6/L9/L10 几何（轴对称 imposed）；8 起源归因（星暴 vs 历史 AGN vs 卫星交互，现有数据不可分）。astra 额外发现合并似然代码缺陷（e079_fit.py:220 对每个 (ON, cam, OFF) 对加一次 OFF 区 deviance —— pooled 拟合 BG 计数 4 次/相机），已由 Claude 读代码确认。本手册已应用的措辞：L7"合并 bin"（§3.4）；L8"模板依赖上限"（§3.5）；L9"轴对称模型内的偏好"（§3.6）；L11 pooled 区间"未校准"（§3.8）；L13"壁归一化误差、斜率模型依赖"（§3.10）；L15"Mach 推定、能量下限条件性"（§4.2）；L14"观测到的现今风相位不能供给"（§3.11 措辞按此精神）。待办：E098 行 61 与 I021 L15 的记录级修改（A1 措辞）；页 v10 面板 P 数字。
R1 (the sol main report plus the partial astra report, two independent arms) ranked the eight weakest links by severity, with per-claim wording downgrades. The sol ranking: 1 L7 endpoint (a pooled bin, not an endpoint; PN alone is weak); 2 L12 halo (the brightness decline is measured, but the absolute density/pressure needs spherical shape, a thermal spectrum, abundances and a fitted constant); 3 L11 O/Fe (a fitted ratio is not absolute enrichment; pooled solar is allowed at 2σ); 4 L13 2.95 ± 0.16 (wall-only error; slope model-dependent); 5 L15 Mach (inferred, not measured — fatal to a unique shock history); 6 L14–L17 (the thin shell omits shell heat; injection power conflated with lifetime-averaged power); 7 L6/L9/L10 geometry (axisymmetry imposed); 8 causal attribution (starburst vs historical AGN vs satellite interaction, not separable with these data). The astra arm additionally found the pooled-likelihood code defect (e079_fit.py:220 adds the OFF-region deviance once per (ON, cam, OFF) pair — a pooled fit counts the BG data 4 times per camera), confirmed by Claude reading the code. Wording already applied in this handbook: L7 "pooled bin" (§3.4); L8 "template-dependent limits" (§3.5); L9 "a preference within axisymmetric families" (§3.6); L11 pooled intervals "uncalibrated" (§3.8); L13 "wall-normalisation error, slope model-dependent" (§3.10); L15 "Mach inferred, energy floor conditional" (§4.2); L14 "the observed present-wind phase cannot supply" (§3.11 wording in that spirit). To do: the record-level fixes of E098 line 61 and I021 L15 (the A1 wording); the page v10 panel P numbers.
来源Source: R1 REPORT.sol.md (8 ranks, wording table); R1 REPORT.astra.md (partial, code defect); session 2026-10-15.md
下表把 L1–L21 按三层归类：就绪（MEASURED 且门已过）、过某个指名门后就绪、仅讨论级/暂缓。每行给出 L 编号与门。所有结果的能量类数字按 ∝ P_halo(47 kpc)/4.1e-14 缩放；māsses 在固定波段发射率下与 P_halo 无关。
The table sorts L1–L21 into three tiers: ready (MEASURED and the gate has passed), ready after a named gate, and discussion-only / not yet. Each row cites its L-number and its gate. All energy-like numbers scale as ∝ P_halo(47 kpc)/4.1e-14; masses are independent of P_halo at fixed band emissivity.
| 层Tier | 主张（L 编号，门）Claim (L-number, gate) | 条件/数字Condition/number |
|---|---|---|
就绪（论文 I 核心）Ready (Paper I core)MEASURED |
L6 X 形双极风（C1 最终图集后可定稿）L6 X-shaped bipolar wind (finalise after the C1 map set) | 逐历元/逐仪器复现reproduced per epoch and per instrument |
| L7 壁到 66 kpc（C1、C3 端点系统学）L7 walls to 66 kpc (C1, C3 extent systematics) | 2.48 ± 0.43 (5.8σ); 2.09 ± 0.67 (3.1σ) | |
| L8 无帽（C4 look-elsewhere）L8 no cap (C4 look-elsewhere) | K95 1.4 / 2.1 PW | |
| L9 几何（C3；轴对称措辞）L9 geometry (C3; axisymmetric wording) | 46° cone → chimney 36–41 kpc | |
| L10 kT(z)（C2 已过；C5 前景/参考区）L10 kT(z) (C2 done; C5 foreground/reference regions) | 0.25–0.31 → 0.19–0.22 keV | |
| L12 晕（C7 FXT 渐晕检验）L12 halo (C7 FXT vignetting test) | nH(47) 5.5e-5, 2σ ≤ 8.6e-5 | |
| 过指名门后Ready after a named gate | L11 O/Fe 径向（论文 II；C5、C6，合并似然修复）L11 radial O/Fe (Paper II; C5, C6, pooled-likelihood fix) | 1T 5–10 相对值；pooled 未校准1T 5–10 relative; pooled uncalibrated |
| L13 超压（C7；误差口径按 R1 措辞）L13 over-pressure (C7; error wording per R1) | 2.95 → 2.87 (C2) | |
| L14/L17 收支（C2 已过；随 L7/L13 更新）L14/L17 budgets (C2 done; update with L7/L13) | ×10–40 现今风×10–40 the present wind | |
| L18 鞘界（需 ~2× 灵敏度或堆叠谱）L18 sheath bounds (need ~2× sensitivity or a stacked spectrum) | 1.36–10× 预测1.36–10× the prediction | |
| 仅讨论级 / 暂缓Discussion only / not yet | L15 弱激波遗迹（C11 理论一致性；Mach 推定）L15 weak-shock relic (C11 theory consistency; Mach inferred) | E0 3.40–4.94e57 erg/lobe ∝ P_halo |
| L16 隐藏晕上限（条件于 f_w）L16 hidden-halo bound (conditional on f_w) | P_tot(47) ≲ 3 P_X | |
| L20/L21 候选（C4/C12 后再评估）L20/L21 candidates (revisit after C4/C12) | 110 kpc 超出；轴 C +2.8 ± 1.0 (PN)110 kpc excess; axis C +2.8 ± 1.0 (PN) | |
| P1–P5、L1、L2、A1 物理讨论（2026-10-15 报告）P1–P5, L1, L2, A1 physics discussions (the 2026-10-15 reports) | 讨论级；见 §4.1–§4.6discussion level; see §4.1–§4.6 |
来源Source: I021_s09_paper_plan_v2.md (L1–L21, C1–C12, papers); 2026-10-15 session + reports (C2, P1–P5, A1, R1, L1, L2)