diff --git a/.custodian-brief.md b/.custodian-brief.md
index 2784301..e2f9f2a 100644
--- a/.custodian-brief.md
+++ b/.custodian-brief.md
@@ -2,16 +2,15 @@
 # Custodian Brief — assembly-economics
 
 **Domain:** communication  
-**Last synced:** 2026-10-12 05:33 UTC  
+**Last synced:** 2026-10-12 05:35 UTC  
 **State Hub:** http://127.0.0.1:8000 *(adjust if running on a remote machine)*
 
 ## Active Workstreams
 
 ### GADM-0.3: global sensitivity, power channels, cohorts
-Progress: 4/6 done  |  workplan_id: `23e33ee7-cb27-52bc-819c-704f44f91663`
+Progress: 5/6 done  |  workplan_id: `23e33ee7-cb27-52bc-819c-704f44f91663`
 
 **Open tasks:**
-- · Strengthen attention dynamics and the full organism grid  `f56e20b4`
 - · Life-cycle cohorts  `2b749127`
 
 ## Inbox Hygiene
diff --git a/AGENTS.md b/AGENTS.md
index 04fa022..2d39e8e 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -23,6 +23,7 @@ cd simulations/gadm02 && python3 gsa.py   # global sensitivity, ~5 min
 cd simulations/gadm02 && python3 -B power_channels.py   # GADM-0.3 power channels (T03 stage)
 cd simulations/gadm02 && python3 -B wealth.py   # GADM-0.3 wealth vs. power
 cd simulations/gadm02 && python3 -B organism_grid.py   # GADM-0.3 organism grid
+cd simulations/gadm02 && python3 -B life_cycle.py   # GADM-0.3 cohorts
 ```
 
 Use `python3 -B` after quick edits; a stale bytecode cache once masked a change.
diff --git a/WORK-RECORDS.md b/WORK-RECORDS.md
index dbfc919..b60a4df 100644
--- a/WORK-RECORDS.md
+++ b/WORK-RECORDS.md
@@ -44,5 +44,5 @@
 | task | ASSEMBLY-WP-0005-T02 | done | — | workplans/ASSEMBLY-WP-0005-gadm-0.3.md |
 | task | ASSEMBLY-WP-0005-T03 | done | — | workplans/ASSEMBLY-WP-0005-gadm-0.3.md |
 | task | ASSEMBLY-WP-0005-T04 | done | — | workplans/ASSEMBLY-WP-0005-gadm-0.3.md |
-| task | ASSEMBLY-WP-0005-T05 | todo | — | workplans/ASSEMBLY-WP-0005-gadm-0.3.md |
+| task | ASSEMBLY-WP-0005-T05 | done | — | workplans/ASSEMBLY-WP-0005-gadm-0.3.md |
 | task | ASSEMBLY-WP-0005-T06 | todo | — | workplans/ASSEMBLY-WP-0005-gadm-0.3.md |
diff --git a/models/GADM-0.3.md b/models/GADM-0.3.md
index 484bf26..739479b 100644
--- a/models/GADM-0.3.md
+++ b/models/GADM-0.3.md
@@ -6,10 +6,10 @@ the code is the reference if the two ever disagree.
 
 # Generative Assembly Dynamics Model — GADM-0.3
 
-**Status:** in progress, exploratory. Implemented so far: power channels
-(§ 2, ASSEMBLY-WP-0005-T03), wealth separated from power (§ 3, T04),
-attention recalibration and the organism grid (§ 4, T05). Planned: cohorts
-(T06).
+**Status:** complete for ASSEMBLY-WP-0005, exploratory. Power channels
+(§ 2, T03), wealth separated from power (§ 3, T04), attention recalibration
+and organism grid (§ 4, T05), life-cycle cohorts (§ 5, T06). Open items are
+in § 6 and ASSEMBLY-WP-0006.
 
 GADM-0.3 builds on GADM-0.2 (`GADM-0.2.md`). Everything not described here is
 unchanged; the implementation reuses `gadm02.py`. Time unit: one year.
@@ -343,6 +343,106 @@ background reproduces the generative regime exactly (VFAC 0.672).
    stays at about 0.6, because it depends on discovery concentration, which
    is not an organism lever.
 
+---
+
+## 5. Life-cycle cohorts (T06)
+
+Implemented in `simulations/gadm02/gadm03c.py` on top of the full GADM-0.3.
+
+### 5.1 Structure
+
+| Cohort | Ages | Carries | Notes |
+|---|---|---|---|
+| children `ch` | 0–19 | H | education, adult attention as care input, seniors' mentoring |
+| working age `w` | 20–64 | H, A | GADM-0.2 learning; deskilling; mentoring inflow T_g; main care provider |
+| seniors `s` | 65–79 | H, A | half learning rate, 1.5× depreciation; provides a share of care |
+| very old `d` | 80+ | H | 3× depreciation; main care need |
+
+Population counts follow births (`birth` × N_w) and ageing rates of 1/20,
+1/45 and 1/15 per year, with deaths of the very old at 1/8 per year. At
+`birth` = 1/45 (replacement), the age structure is stationary at about
+23 / 51 / 17 / 9 %. A unit test checks this.
+
+Aggregates passed to the rest of the model: H is population-weighted over all
+cohorts; A is weighted over working age and seniors. G remains a single
+aggregate.
+
+Care need = 0.5 × share of children + 1.0 × share of very old. Its attention
+cost is split between working age and seniors (`care_sen`), and reduced by
+participation claims: relief = 1 − 3 × cp.
+
+### 5.2 Effect on the scenarios
+
+With cohorts, the baseline VFAC is 0.323 (0.334 without cohorts) and the
+generative regime 0.656 (0.672). Capability and attention are a little lower,
+because of senior depreciation and care load. The scenario ordering is
+unchanged.
+
+### 5.3 Hypotheses L1–L3
+
+Full tables: `simulations/gadm02/results/life_cycle.md`.
+
+**L1: strong intergenerational transfer lets an ageing society keep H.
+Supported.**
+
+| Demography | T_g = 0 | T_g = 0.05 | T_g = 0.15 |
+|---|---|---|---|
+| stable (ΔH over 100 years) | −0.043 | −0.032 | −0.011 |
+| ageing, births at 75 % of replacement | −0.053 | −0.040 | −0.018 |
+
+An ageing society with strong mentoring loses less capability than a stable
+society without it. A first formulation of T_g, which transferred only
+max(0, H_s − H_w), was inert because senior capability decays below
+working-age capability. It was replaced by an inflow proportional to senior
+capability and working-age headroom.
+
+**L2: without recognized care participation, A falls fastest in the
+working-age cohort. Conditionally supported.** Attention is lowest in
+whichever cohort carries the higher care load *per person*:
+
+- with seniors providing 10 % of care, working-age A is lowest (0.40 vs. 0.44);
+- with 30 %, senior A is lowest (0.40 vs. 0.42), because there are fewer
+  seniors to share it.
+
+Care relief through participation claims (cp = 0.15) raises working-age
+attention by 0.034–0.042 in every setting. So the post-work participation
+channel directly protects the attention of carers.
+
+**L3: education reaches working-age capability with a 15–20 year lag. Not
+supported in this model.** Raising education intensity by 50 % raises
+children's capability by +0.067, but working-age capability only by +0.010,
+with half of the effect reached after about 9 years. The cause is structural:
+working-age capability relaxes to its own conditions within roughly 8 years
+(adult learning and depreciation rates). That washes out the cohort inflow,
+so H has almost no memory of childhood. Human-capital research generally
+finds persistent effects of early education, so this points to a missing
+mechanism, not an empirical result. The fix belongs in GADM-0.4: split H
+into a durable, education-formed component and a current component.
+
+### 5.4 Further observations
+
+- **Ageing is slow in this demography.** At 75 % of replacement fertility,
+  the share of seniors and very old rises only from 0.26 to 0.29 in 100 years,
+  and total care need barely changes (fewer children offset more very old at
+  the chosen weights).
+- **Aggregate H declines at baseline even with stationary demography**
+  (−0.043), because of senior and very-old depreciation, which the
+  single-population model did not have.
+
+---
+
+## 6. Open items (handed to ASSEMBLY-WP-0006)
+
+- Split H into durable and current components (L3).
+- Agency G per cohort.
+- Initialise W and the cohort attention stocks at their own equilibria (§ 3.4
+  reading 3, § 5.3 L2 note).
+- Estimates instead of assumptions for the channel self-reinforcement r_k, the
+  wealth conversion k_w, and organism depreciation `dO` (the top driver in the
+  GADM-0.2 Sobol analysis).
+- Global sensitivity analysis on GADM-0.3 (the Sobol run so far covers
+  GADM-0.2).
+
 ## Reproduce
 
 ```bash
@@ -350,4 +450,5 @@ python3 -B -m unittest discover -s simulations/gadm02
 cd simulations/gadm02 && python3 -B power_channels.py   # T03 stage (wealth off)
 cd simulations/gadm02 && python3 -B wealth.py           # T04: P3 and scenarios
 cd simulations/gadm02 && python3 -B organism_grid.py    # T05: 2 × 2 × 2 organism grid
+cd simulations/gadm02 && python3 -B life_cycle.py       # T06: cohorts, L1–L3
 ```
diff --git a/models/README.md b/models/README.md
index 0e7657d..2becddd 100644
--- a/models/README.md
+++ b/models/README.md
@@ -11,7 +11,7 @@ exploratory: they generate hypotheses and are not established results.
 | [InformationOrganismsInGADM.md](InformationOrganismsInGADM.md) | mapping note v0.1 | ASSEMBLY-WP-0003-T03; input for GADM-0.2 (ASSEMBLY-WP-0004) |
 | [GADM-CLD.md](GADM-CLD.md) | v0.1 | causal-loop diagram, loops, signed link table (ASSEMBLY-WP-0004-T01/T02) |
 | [GADM-0.2.md](GADM-0.2.md) | GADM-0.2 | extensions decision, equations, scenario and sensitivity findings (ASSEMBLY-WP-0004) |
-| [GADM-0.3.md](GADM-0.3.md) | in progress | power channels (P1/P2) wealth separated from power (P3), attention recalibration and organism grid (ASSEMBLY-WP-0005-T03–T05) |
+| [GADM-0.3.md](GADM-0.3.md) | GADM-0.3 | power channels (P1/P2) wealth separated from power (P3), attention recalibration, organism grid, life-cycle cohorts (ASSEMBLY-WP-0005) |
 | [GADM-0.3-Candidates.md](GADM-0.3-Candidates.md) | outline | life-cycle and power extensions, deferred items (ASSEMBLY-WP-0004-T08) |
 
 ## Conventions
diff --git a/simulations/README.md b/simulations/README.md
index 82cdde1..f116b84 100644
--- a/simulations/README.md
+++ b/simulations/README.md
@@ -6,7 +6,8 @@ model version it implements.
 | Directory | Model | Entry points |
 |-----------|-------|--------------|
 | `gadm02/` | GADM-0.2 (`models/GADM-0.2.md`) | `run_scenarios.py`, `sensitivity.py`, `gsa.py`, `test_gadm02.py` |
-| `gadm02/gadm03.py` | GADM-0.3: power channels, wealth (`models/GADM-0.3.md`) | `power_channels.py`, `wealth.py`, `organism_grid.py`, `test_gadm03.py` |
+| `gadm02/gadm03.py` | GADM-0.3: power channels, wealth, attention (`models/GADM-0.3.md`) | `power_channels.py`, `wealth.py`, `organism_grid.py`, `test_gadm03.py` |
+| `gadm02/gadm03c.py` | GADM-0.3 with life-cycle cohorts (`models/GADM-0.3.md` § 5) | `life_cycle.py`, `test_gadm03c.py` |
 
 Python 3.10+, standard library only. Use `python3 -B` after quick edits: a stale
 bytecode cache can mask same-size changes made within the same second.
diff --git a/simulations/gadm02/gadm03c.py b/simulations/gadm02/gadm03c.py
new file mode 100644
index 0000000..1d2dd38
--- /dev/null
+++ b/simulations/gadm02/gadm03c.py
@@ -0,0 +1,141 @@
+"""GADM-0.3 with life-cycle cohorts (ASSEMBLY-WP-0005-T06).
+
+Extends gadm03 (power channels, wealth, recalibrated attention). Human
+capability H and attention headroom A become cohort quantities; everything else
+is unchanged and sees population-weighted aggregates:
+
+    cohorts: ch (children, 0-19), w (working age, 20-64),
+             s (seniors, 65-79), d (very old / dependent, 80+)
+
+Population counts N_c evolve by births (b * N_w), ageing (1/20, 1/45, 1/15 per
+year) and deaths of the very old (1/8 per year). With b = 1/45 the age
+structure is stationary (shares ≈ 23 / 51 / 17 / 9 %).
+
+Intensive cohort quantities follow their own dynamics plus a mixing term when
+people age in: dX_c += (inflow_c / N_c) * (X_from - X_c).
+
+Effective agency G stays a single aggregate in this version.
+
+Mentoring (T_g) is an inflow proportional to seniors' capability and
+working-age headroom. A first version transferred only max(0, Hs - Hw); since
+seniors' capability decays below working-age capability, that term was inert
+(L1 showed zero effect) and was replaced.
+"""
+
+from __future__ import annotations
+
+import gadm02 as g2
+import gadm03 as g3
+
+MODEL_VERSION = "GADM-0.3-cohorts"
+TIME_UNIT = g3.TIME_UNIT
+
+COHORTS = ("ch", "w", "s", "d")
+AGE_RATE = {"ch": 1 / 20, "w": 1 / 45, "s": 1 / 15}
+DEATH_D = 1 / 8
+
+STOCKS = tuple(k for k in g3.STOCKS if k not in ("H", "A")) + (
+    "Nch", "Nw", "Ns", "Nd", "Hch", "Hw", "Hs", "Hd", "Aw", "As")
+
+INIT = {k: v for k, v in g3.INIT.items() if k not in ("H", "A")}
+INIT.update({
+    # stationary age structure for b = 1/45 (shares of a population of 1)
+    "Nch": 0.2273, "Nw": 0.5114, "Ns": 0.1705, "Nd": 0.0909,
+    "Hch": 0.40, "Hw": 0.55, "Hs": 0.55, "Hd": 0.40,   # aggregate H ≈ 0.50
+    "Aw": 0.45, "As": 0.65,                             # aggregate adult A ≈ 0.50
+})
+
+PARAMS = dict(g3.PARAMS)
+PARAMS.update({
+    "birth": 1 / 45,   # births per working-age person per year (1/45 = replacement)
+    "edu": 1.0,        # education intensity for children (multiplier)
+    "tg": 0.05,        # intergenerational transfer: seniors' mentoring of working age
+    "tgc": 0.03,       # seniors' mentoring of children
+    "dHs": 1.5,        # capability depreciation multiplier, seniors
+    "dHd": 3.0,        # capability depreciation multiplier, very old
+    "learn_s": 0.5,    # learning multiplier for seniors
+    "care_load": 0.12, # attention cost of care per unit of care need
+    "care_sen": 0.3,   # share of informal care provided by seniors
+    "care_relief": 3.0,  # relief of informal care per unit of participation claims cp
+})
+
+
+def shares(s: dict) -> dict:
+    tot = s["Nch"] + s["Nw"] + s["Ns"] + s["Nd"]
+    return {c: s["N" + c] / tot for c in COHORTS}
+
+
+def aggregates(s: dict) -> dict:
+    n = shares(s)
+    H = sum(n[c] * s["H" + c] for c in COHORTS)
+    A = (n["w"] * s["Aw"] + n["s"] * s["As"]) / (n["w"] + n["s"])
+    return {"H": H, "A": A, **{"n_" + c: n[c] for c in COHORTS}}
+
+
+def derivatives(s: dict, p: dict) -> tuple[dict, dict]:
+    agg = aggregates(s)
+    base = {k: v for k, v in s.items() if k in g3.STOCKS}
+    base["H"], base["A"] = agg["H"], agg["A"]
+    d3, a = g3.derivatives(base, p)
+    d = {k: v for k, v in d3.items() if k not in ("H", "A")}
+
+    S, B, X, Ko, m = s["S"], a["B"], s["X"], s["Ko"], p["m"]
+    n = shares(s)
+    Nch, Nw, Ns, Nd = s["Nch"], s["Nw"], s["Ns"], s["Nd"]
+
+    # demography
+    births = p["birth"] * Nw
+    flow = {"ch": AGE_RATE["ch"] * Nch, "w": AGE_RATE["w"] * Nw, "s": AGE_RATE["s"] * Ns}
+    d["Nch"] = births - flow["ch"]
+    d["Nw"] = flow["ch"] - flow["w"]
+    d["Ns"] = flow["w"] - flow["s"]
+    d["Nd"] = flow["s"] - DEATH_D * Nd
+
+    # human capability per cohort (GADM-0.2 form, cohort-specific modifiers)
+    social = p["h0"] * (0.5 * S + 0.5 * B)
+    desk = p["desk"] * Ko * (1 - m)
+    Hch, Hw, Hs, Hd = s["Hch"], s["Hw"], s["Hs"], s["Hd"]
+    d["Hch"] = (p["edu"] * (social + p["hT"] * X) + p["hA"] * s["Aw"] + p["tgc"] * n["s"] / max(n["ch"], 1e-9) * Hs) \
+        * (1 - Hch) - p["dH"] * Hch
+    d["Hw"] = (social + p["hA"] * s["Aw"] + p["hT"] * X) * (1 - Hw) - (p["dH"] + desk) * Hw \
+        + p["tg"] * n["s"] / max(n["w"], 1e-9) * Hs * (1 - Hw) \
+        + flow["ch"] / Nw * (Hch - Hw)
+    d["Hs"] = p["learn_s"] * (social + p["hA"] * s["As"] + p["hT"] * X) * (1 - Hs) \
+        - (p["dHs"] * p["dH"] + desk) * Hs + flow["w"] / Ns * (Hw - Hs)
+    d["Hd"] = -p["dHd"] * p["dH"] * Hd + flow["s"] / Nd * (Hs - Hd)
+
+    # attention headroom: extraction (as GADM-0.3) plus informal care load
+    care_need = 0.5 * n["ch"] + 1.0 * n["d"]
+    relief = max(0.0, 1 - p["care_relief"] * p["cp"])
+    load = p["care_load"] * care_need * relief
+    ext = p["aext"] * p["x_att"] * (0.2 + Ko)
+    rec = p["arec"] * (0.4 + 0.6 * B)
+    d["Aw"] = rec * (1 - s["Aw"]) - (ext + load * (1 - p["care_sen"]) / n["w"]) * s["Aw"] \
+        + flow["ch"] / Nw * (0.5 - s["Aw"])       # young adults enter with A = 0.5
+    d["As"] = rec * (1 - s["As"]) - (ext + load * p["care_sen"] / n["s"]) * s["As"] \
+        + flow["w"] / Ns * (s["Aw"] - s["As"])
+    a = {**a, **agg, "care_need": care_need}
+    return d, a
+
+
+def simulate(params: dict | None = None, init: dict | None = None,
+             T: float = 100.0, dt: float = 0.1, record_every: int = 10) -> list[dict]:
+    p = {**PARAMS, **(params or {})}
+    s = {**INIT, **(init or {})}
+    unknown = set(p) - set(PARAMS)
+    if unknown:
+        raise KeyError(f"unknown parameters: {sorted(unknown)}")
+    rows = []
+    steps = int(round(T / dt))
+    for i in range(steps + 1):
+        d, a = derivatives(s, p)
+        if i % record_every == 0:
+            row = {"t": round(i * dt, 6), **s, "P": g3.total_p(s),
+                   **{k: a[k] for k in ("H", "A", "n_ch", "n_w", "n_s", "n_d", "care_need",
+                                        "Ystar", "B", "V", "Vsel")}}
+            row["VFAC"] = g2.vfac(row)
+            rows.append(row)
+        if i < steps:
+            s = {k: (max(0.0, s[k] + dt * d[k]) if k.startswith("N")
+                     else min(1.0, max(0.0, s[k] + dt * d[k]))) for k in STOCKS}
+    return rows
diff --git a/simulations/gadm02/life_cycle.py b/simulations/gadm02/life_cycle.py
new file mode 100644
index 0000000..6ec277b
--- /dev/null
+++ b/simulations/gadm02/life_cycle.py
@@ -0,0 +1,80 @@
+"""Life-cycle hypotheses L1-L3 for GADM-0.3 with cohorts (gadm03c).
+
+L1: strong intergenerational transfer T_g lets an ageing society keep H.
+L2: care demand competes with attention; without recognized care participation
+    (C_p), attention headroom falls fastest in the working-age cohort.
+L3: children's capability formation reaches working-age capability with a lag
+    of 15-20 years.
+
+Usage: python3 -B simulations/gadm02/life_cycle.py
+"""
+
+from __future__ import annotations
+
+import sys
+from pathlib import Path
+
+sys.dont_write_bytecode = True
+
+import gadm03c as c  # noqa: E402
+from scenarios import load_all  # noqa: E402
+
+OUT = Path(__file__).resolve().parent / "results"
+STABLE, AGEING = 1 / 45, 0.75 / 45
+
+
+def run(params: dict, base: dict) -> list[dict]:
+    return c.simulate({**base["params"], **params}, base["init"])
+
+
+def main() -> None:
+    OUT.mkdir(exist_ok=True)
+    sc = load_all()["baseline"]
+    lines = [f"# {c.MODEL_VERSION}: life-cycle hypotheses", "",
+             "Generated by `simulations/gadm02/life_cycle.py`. Baseline background,",
+             "t = 100 years. Ageing demography: births at 75 % of replacement.", ""]
+
+    lines += ["## L1: intergenerational transfer in an ageing society", "",
+              "| Demography | T_g (tg) | seniors + very old (share, t=100) | H (t=0) | H (t=100) | ΔH |",
+              "|---|---|---|---|---|---|"]
+    for demo, b in (("stable", STABLE), ("ageing", AGEING)):
+        for tg in (0.0, 0.05, 0.15):
+            r = run({"birth": b, "tg": tg}, sc)
+            old = r[-1]["n_s"] + r[-1]["n_d"]
+            lines.append(f"| {demo} | {tg:.2f} | {old:.2f} | {r[0]['H']:.3f} | {r[-1]['H']:.3f} | "
+                         f"{r[-1]['H'] - r[0]['H']:+.3f} |")
+
+    lines += ["", "## L2: care load and attention", "",
+              "Levels at t = 100 are compared across settings (not against t = 0, because",
+              "initial cohort attention is not at its own equilibrium).", "",
+              "| Demography | care_sen (seniors' care share) | cp | A working age | A seniors | care load per working-age person | per senior |",
+              "|---|---|---|---|---|---|---|"]
+    for demo, b in (("stable", STABLE), ("ageing", AGEING)):
+        for cs in (0.1, 0.3):
+            for cp in (0.0, 0.15):
+                prm = {"birth": b, "cp": cp, "care_sen": cs}
+                r = run(prm, sc)[-1]
+                pp = {**c.PARAMS, **sc["params"], **prm}
+                load = pp["care_load"] * r["care_need"] * max(0.0, 1 - pp["care_relief"] * cp)
+                lines.append(f"| {demo} | {cs:.1f} | {cp:.2f} | {r['Aw']:.3f} | {r['As']:.3f} | "
+                             f"{load * (1 - cs) / r['n_w']:.4f} | {load * cs / r['n_s']:.4f} |")
+
+    lines += ["", "## L3: lag from education to working-age capability", "",
+              "Run difference: education intensity 1.5 vs. 1.0, everything else equal.", "",
+              "| t | ΔH children | ΔH working age | ΔH aggregate |", "|---|---|---|---|"]
+    lo, hi = run({"edu": 1.0}, sc), run({"edu": 1.5}, sc)
+    diffs = [(a["t"], b["Hch"] - a["Hch"], b["Hw"] - a["Hw"], b["H"] - a["H"]) for a, b in zip(lo, hi)]
+    for t, dc, dw, dh in diffs:
+        if t in (0, 5, 10, 15, 20, 25, 30, 40, 50, 75, 100):
+            lines.append(f"| {t:.0f} | {dc:+.3f} | {dw:+.3f} | {dh:+.3f} |")
+    final_w, final_c = diffs[-1][2], diffs[-1][1]
+    t_w = next(t for t, _dc, dw, _dh in diffs if dw >= 0.5 * final_w)
+    t_c = next(t for t, dc, _dw, _dh in diffs if dc >= 0.5 * final_c)
+    lines += ["", f"Half of the final effect is reached after {t_c:.0f} years for children and "
+              f"{t_w:.0f} years for working-age capability.", ""]
+    (OUT / "life_cycle.md").write_text("\n".join(lines))
+    print("\n".join(lines))
+
+
+if __name__ == "__main__":
+    main()
diff --git a/simulations/gadm02/results/life_cycle.md b/simulations/gadm02/results/life_cycle.md
new file mode 100644
index 0000000..80e3f25
--- /dev/null
+++ b/simulations/gadm02/results/life_cycle.md
@@ -0,0 +1,51 @@
+# GADM-0.3-cohorts: life-cycle hypotheses
+
+Generated by `simulations/gadm02/life_cycle.py`. Baseline background,
+t = 100 years. Ageing demography: births at 75 % of replacement.
+
+## L1: intergenerational transfer in an ageing society
+
+| Demography | T_g (tg) | seniors + very old (share, t=100) | H (t=0) | H (t=100) | ΔH |
+|---|---|---|---|---|---|
+| stable | 0.00 | 0.26 | 0.502 | 0.459 | -0.043 |
+| stable | 0.05 | 0.26 | 0.502 | 0.470 | -0.032 |
+| stable | 0.15 | 0.26 | 0.502 | 0.491 | -0.011 |
+| ageing | 0.00 | 0.29 | 0.502 | 0.450 | -0.053 |
+| ageing | 0.05 | 0.29 | 0.502 | 0.462 | -0.040 |
+| ageing | 0.15 | 0.29 | 0.502 | 0.485 | -0.018 |
+
+## L2: care load and attention
+
+Levels at t = 100 are compared across settings (not against t = 0, because
+initial cohort attention is not at its own equilibrium).
+
+| Demography | care_sen (seniors' care share) | cp | A working age | A seniors | care load per working-age person | per senior |
+|---|---|---|---|---|---|---|
+| stable | 0.1 | 0.00 | 0.399 | 0.438 | 0.0432 | 0.0144 |
+| stable | 0.1 | 0.15 | 0.441 | 0.465 | 0.0238 | 0.0079 |
+| stable | 0.3 | 0.00 | 0.417 | 0.395 | 0.0336 | 0.0432 |
+| stable | 0.3 | 0.15 | 0.452 | 0.439 | 0.0185 | 0.0238 |
+| ageing | 0.1 | 0.00 | 0.402 | 0.443 | 0.0407 | 0.0128 |
+| ageing | 0.1 | 0.15 | 0.443 | 0.468 | 0.0224 | 0.0070 |
+| ageing | 0.3 | 0.00 | 0.420 | 0.403 | 0.0317 | 0.0383 |
+| ageing | 0.3 | 0.15 | 0.454 | 0.444 | 0.0174 | 0.0210 |
+
+## L3: lag from education to working-age capability
+
+Run difference: education intensity 1.5 vs. 1.0, everything else equal.
+
+| t | ΔH children | ΔH working age | ΔH aggregate |
+|---|---|---|---|
+| 0 | +0.000 | +0.000 | +0.000 |
+| 5 | +0.053 | +0.003 | +0.013 |
+| 10 | +0.067 | +0.006 | +0.018 |
+| 15 | +0.069 | +0.008 | +0.020 |
+| 20 | +0.069 | +0.008 | +0.020 |
+| 25 | +0.068 | +0.009 | +0.020 |
+| 30 | +0.068 | +0.009 | +0.020 |
+| 40 | +0.067 | +0.009 | +0.020 |
+| 50 | +0.067 | +0.009 | +0.020 |
+| 75 | +0.067 | +0.010 | +0.021 |
+| 100 | +0.067 | +0.010 | +0.021 |
+
+Half of the final effect is reached after 3 years for children and 9 years for working-age capability.
diff --git a/simulations/gadm02/test_gadm03c.py b/simulations/gadm02/test_gadm03c.py
new file mode 100644
index 0000000..7c69805
--- /dev/null
+++ b/simulations/gadm02/test_gadm03c.py
@@ -0,0 +1,44 @@
+"""Sanity tests for GADM-0.3 with cohorts.
+
+Usage: python3 -B -m unittest discover -s simulations/gadm02
+"""
+
+import unittest
+
+import gadm03c as c
+from scenarios import load_all
+
+
+class TestCohorts(unittest.TestCase):
+    def test_bounded_and_shares_sum_to_one(self):
+        for name, sc in load_all().items():
+            for row in c.simulate(sc["params"], sc["init"]):
+                for k in c.STOCKS:
+                    if not k.startswith("N"):
+                        self.assertGreaterEqual(row[k], 0.0, (name, k))
+                        self.assertLessEqual(row[k], 1.0, (name, k))
+                self.assertAlmostEqual(row["n_ch"] + row["n_w"] + row["n_s"] + row["n_d"], 1.0)
+
+    def test_replacement_fertility_is_stationary(self):
+        r = c.simulate()
+        for k in ("n_ch", "n_w", "n_s", "n_d"):
+            self.assertAlmostEqual(r[0][k], r[-1][k], places=3, msg=k)
+
+    def test_low_fertility_ages_the_population(self):
+        r = c.simulate({"birth": 0.75 / 45})[-1]
+        s = c.simulate()[-1]
+        self.assertGreater(r["n_s"] + r["n_d"], s["n_s"] + s["n_d"])
+
+    def test_mentoring_raises_capability(self):
+        lo = c.simulate({"tg": 0.0})[-1]["H"]
+        hi = c.simulate({"tg": 0.15})[-1]["H"]
+        self.assertGreater(hi, lo)
+
+    def test_care_relief_raises_working_age_attention(self):
+        lo = c.simulate({"cp": 0.0})[-1]["Aw"]
+        hi = c.simulate({"cp": 0.15})[-1]["Aw"]
+        self.assertGreater(hi, lo)
+
+
+if __name__ == "__main__":
+    unittest.main()
diff --git a/workplans/ASSEMBLY-WP-0005-gadm-0.3.md b/workplans/ASSEMBLY-WP-0005-gadm-0.3.md
index fd9473e..8e37468 100644
--- a/workplans/ASSEMBLY-WP-0005-gadm-0.3.md
+++ b/workplans/ASSEMBLY-WP-0005-gadm-0.3.md
@@ -4,7 +4,7 @@ type: workplan
 title: "GADM-0.3: global sensitivity, power channels, cohorts"
 domain: communication
 repo: assembly-economics
-status: active
+status: finished
 flavor: residual
 origin: residual
 origin_ref: ASSEMBLY-WP-0004
@@ -89,10 +89,32 @@ organism scenario grid (lineage × membrane × attention regime).
 
 ```task
 id: ASSEMBLY-WP-0005-T06
-status: todo
+status: done
 priority: low
 state_hub_task_id: "2b749127-9f41-5bff-9a56-19ccd8fe9dd6"
 ```
 
 Implement `GADM-0.3-Candidates.md` § 1: four cohorts, intergenerational
 transfer, and care demand on attention. Test hypotheses L1–L3.
+
+## Outcome
+
+All tasks are done. Results are in `models/GADM-0.3.md` and
+`models/GADM-0.2.md` § 5.1.
+
+- T01: one period = one year; global sensitivity analysis. The model is
+  nearly additive. The "reuse as a bundle" reading was refuted, and organism
+  governance levers were confirmed as dominant.
+- T02: literature verified; AI-growth prior art added; novelty claim 2
+  narrowed.
+- T03: six power channels. P1 is mostly unsupported, except lineage
+  hysteresis; P2 is supported.
+- T04: wealth separated from power. P3 is supported and robust.
+- T05: attention recalibrated against time-use data. In the 2 × 2 × 2 grid,
+  lineage ownership is the strongest organism lever.
+- T06: life-cycle cohorts. L1 is supported (after fixing an inert mentoring
+  term), L2 is conditional, and L3 is not supported (a missing durable-capability
+  mechanism).
+
+Residual: ASSEMBLY-WP-0006 (GADM-0.4, proposed) carries the open items in
+`GADM-0.3.md` § 6.
diff --git a/workplans/ASSEMBLY-WP-0006-gadm-0.4.md b/workplans/ASSEMBLY-WP-0006-gadm-0.4.md
new file mode 100644
index 0000000..45be6c9
--- /dev/null
+++ b/workplans/ASSEMBLY-WP-0006-gadm-0.4.md
@@ -0,0 +1,91 @@
+---
+id: ASSEMBLY-WP-0006
+type: workplan
+title: "GADM-0.4: durable capability, cohort agency, estimated parameters"
+domain: communication
+repo: assembly-economics
+status: proposed
+flavor: residual
+origin: residual
+origin_ref: ASSEMBLY-WP-0005
+depends_on: [ASSEMBLY-WP-0005]
+owner: claude-code
+topic_slug: coulomb-social
+created: "2026-10-12"
+updated: "2026-10-12"
+---
+
+# GADM-0.4: durable capability, cohort agency, estimated parameters
+
+Residual of ASSEMBLY-WP-0005: the open items in `models/GADM-0.3.md` § 6, plus
+the GADM-0.1 § 26 plan for GADM-0.4. Not to be started until it is promoted
+to `ready`.
+
+## Split human capability into durable and current components
+
+```task
+id: ASSEMBLY-WP-0006-T01
+status: todo
+priority: high
+```
+
+L3 failed because H has no memory of childhood. Add a slowly depreciating,
+education-formed component next to the current one, then re-test L3.
+
+## Ground key parameters
+
+```task
+id: ASSEMBLY-WP-0006-T02
+status: todo
+priority: high
+```
+
+Replace the assumed values of organism depreciation `dO` (the top Sobol
+driver), channel self-reinforcement r_k and wealth conversion k_w with
+literature- or data-based ranges. Re-run the affected readings.
+
+## Global sensitivity on GADM-0.3
+
+```task
+id: ASSEMBLY-WP-0006-T03
+status: todo
+priority: medium
+```
+
+Extend `gsa.py` to the GADM-0.3 models (channels, wealth, cohorts). Check
+whether the near-additivity of GADM-0.2 survives the superlinear channel
+dynamics.
+
+## Equilibrium initialisation
+
+```task
+id: ASSEMBLY-WP-0006-T04
+status: todo
+priority: medium
+```
+
+Initialise W and the cohort attention stocks at their own equilibria, to
+remove the initial-condition confounds noted in `GADM-0.3.md` §§ 3.4 and 5.3.
+
+## Agency per cohort
+
+```task
+id: ASSEMBLY-WP-0006-T05
+status: todo
+priority: low
+```
+
+Give G a cohort structure (restartability and exit options differ strongly
+by age).
+
+## Sectors and rival goods
+
+```task
+id: ASSEMBLY-WP-0006-T06
+status: todo
+priority: low
+```
+
+GADM-0.1 § 26 plan for GADM-0.4: multiple sectors, rival versus non-rival
+goods, ecological resource classes, institutional selection mechanisms.
+Scope this into its own workplan before implementing it.