technomaton · jurby · Jun 1, 2026 · Jun 1, 2026
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,5 +1,36 @@
 # Changelog
 
+## 2.1.10 — 2026-06-01 — Drop dead `cadence:` config (cadence is iteration-sourced)
+
+The `cadence:` block (`iteration_weeks` / `pi_weeks` / `delivery_iterations_per_pi`
+/ `ip_iterations_per_pi`) that shipped in the `people.yaml` template was **never read
+by the engine**. Effective cadence is derived per Planning Interval from the `pi:`
+block of `.edpa/iterations/PI-*.yaml` (`_pi_loader.derive_pis`), or inferred from each
+iteration's `weeks:`/dates. The block was dead config that invited "edit this to change
+the rhythm" mistakes, and it had been copied — with a misleading "must mirror" note —
+into `edpa.yaml` examples and test fixtures too.
+
+### refactor(cadence): remove the dead block everywhere, redirect to the live surface
+- `people.yaml.tmpl`: drop the `cadence:` block; add a breadcrumb pointing at the
+  iteration `pi:` block. `people.yaml` is now just `teams` + `people`.
+- Examples/fixtures: drop `cadence:` from `docs/kashealth-pilot/{edpa,people}.yaml.example`,
+  `docs/examples/capacity-small-team.yaml`, and `tests/e2e_v2_full/fixtures/{edpa,people}.yaml`
+  (incl. the stale "must mirror / Wave C verifies this match" comments — no code ever
+  verified it).
+- Docs + web: remove the `cadence:` snippet from README / SETUP / quick-start / playbook
+  (CZ+EN) and the web setup wizard's generated config; point users at the iteration
+  `pi:` block. `docs/cadence.md` gains a "Where cadence is configured" section.
+  Conceptual cadence content (2/10 vs 1/5 variants) is unchanged.
+
+### docs(mcp,backlog): fix two stale "cadence" references
+- `mcp_server.py`: the `edpa://config` resource serves `edpa.yaml` but was described as
+  "Master config: PI, iterations, cadence, sync settings" — corrected to the project
+  config it actually serves.
+- `backlog.py`: the "Load team metadata (people, teams, cadence)" comment overstated —
+  cadence is loaded into the dict but never read; the comment now says so.
+
+No engine behaviour change — cadence was already iteration-sourced. Full suite 577 passing.
+
 ## 2.1.9 — 2026-06-01 — Windows onboarding fixes (filelock, UTF-8 console + file I/O)
 
 `/edpa:edpa-setup` crashed on a fresh Windows box, surfaced by colleagues running

diff --git a/README.md b/README.md
@@ -103,10 +103,6 @@ The template ships with placeholder names. Replace them with your team
 (1-week iterations, 5-week PI = 4 delivery + 1 IP):
 
 ```yaml
-cadence:
-  iteration_weeks: 1        # AI-native default; use 2 for classic SAFe
-  pi_weeks: 5               # 4 delivery iterations + 1 IP
-
 people:
   - id: alice
     name: "Alice Architect"

diff --git a/SETUP.md b/SETUP.md
@@ -34,9 +34,6 @@ If you prefer to configure manually instead of using `/edpa setup`:
 Edit `.edpa/config/people.yaml`:
 
 ```yaml
-cadence:
-  iteration_weeks: 2    # 1 (AI-native) or 2 (classic)
-
 people:
   - id: alice
     name: "Alice Smith"

diff --git a/docs/cadence.md b/docs/cadence.md
@@ -33,3 +33,19 @@ Switch after first PI if:
 ## EDPA is cadence-agnostic
 
 The formula works identically for both variants — only `Capacity[P, I]` changes.
+
+## Where cadence is configured
+
+Cadence is **not** set in `people.yaml`. The engine derives it per Planning
+Interval from the iteration files in `.edpa/iterations/`:
+
+```yaml
+# .edpa/iterations/PI-2026-1.yaml
+pi:
+  id: PI-2026-1
+  iteration_weeks: 1      # 1 (AI-native) or 2 (classic)
+  pi_iterations: 5        # total iterations incl. IP (4 delivery + 1 IP)
+```
+
+If `iteration_weeks` / `pi_iterations` are omitted, the engine infers them from
+each iteration's `weeks:` (or its start/end dates) and the iteration count.
diff --git a/docs/examples/capacity-small-team.yaml b/docs/examples/capacity-small-team.yaml
@@ -1,16 +1,8 @@
-# Example: 3-person startup team
-# Classic SAFe cadence (2-week iterations, 10-week PI).
-#
-# As of 1.4 the default ships AI-native (1-week iter / 5-week PI) — see
-# plugin/edpa/templates/people.yaml.tmpl for the new defaults. This file
-# stays as a reference for teams that want classic SAFe; just halve every
-# capacity_per_iteration value below if you switch to iteration_weeks: 1.
-
-cadence:
-  iteration_weeks: 2
-  pi_weeks: 10
-  delivery_iterations_per_pi: 4
-  ip_iterations_per_pi: 1
+# Example: 3-person startup team.
+# Capacities below assume classic SAFe (2-week iterations): 1.0 FTE = 80h.
+# For the AI-native default (1-week iterations) halve every
+# capacity_per_iteration. Cadence itself is set per-PI in the `pi:` block of
+# .edpa/iterations/PI-*.yaml, not here — see docs/cadence.md.
 
 teams:
   - id: "Acme Corp"

diff --git a/docs/kashealth-pilot/KASHEALTH-PILOT.md b/docs/kashealth-pilot/KASHEALTH-PILOT.md
@@ -315,7 +315,7 @@ Pokud 5+ z 6 PASS → pilot úspěšný, pokračuj na PI-2026-2 (full prod) a zv
 
 ## 7. Open questions (pre-kickoff sync)
 
-1. **PI cadence** — 1-week × 5 (default) vs 2-week × 5? Nastavitelné v `cadence:` (edpa.yaml + people.yaml, musí se shodovat).
+1. **PI cadence** — 1-week × 5 (default) vs 2-week × 5? Nastavitelné per-PI v `pi:` bloku `.edpa/iterations/PI-*.yaml` (`iteration_weeks` / `pi_iterations`).
 2. **FTE distribuce** — 1.0 / 0.5 / 0.25 per člen? Doporučení v `people.yaml.example`.
 3. **Cost reporting** — sazby drží **privátní registr** (ne EDPA people.yaml — engine `hourly_rate` nečte). Auditor format = open question.
 4. **Calibration timing** — `/edpa:autocalib` až po PI-2026-1 close (potřeba ≥ 20 ground truth records).

diff --git a/docs/kashealth-pilot/edpa.yaml.example b/docs/kashealth-pilot/edpa.yaml.example
@@ -58,15 +58,6 @@ governance:
   # audit_mode — snapshoty vždy nesou plný signals[] audit trail.
   methodology: "EDPA 2.1.8"
 
-# Kadence (AI-native default): 1-week iterace × 5 (4 delivery + 1 IP).
-# Musí zrcadlit `cadence:` blok v people.yaml. Pro classic SAFe nastav
-# iteration_weeks: 2 a pi_weeks: 10.
-cadence:
-  iteration_weeks: 1
-  pi_weeks: 5
-  delivery_iterations_per_pi: 4
-  ip_iterations_per_pi: 1
-
 naming:
   pi_pattern: "PI-{year}-{pi_num}"
   iteration_pattern: "PI-{year}-{pi_num}.{iter_num}"

diff --git a/docs/kashealth-pilot/people.yaml.example b/docs/kashealth-pilot/people.yaml.example
@@ -19,15 +19,9 @@
 # ani cost (jen Item, Level, JS, Person, CW, Score, Ratio, Hours).
 # Sazby držte v separátním (privátním) registru — dle smluv —
 # ne v EDPA configu, který může skončit v public repu.
-
-cadence:
-  # AI-native default per EDPA: 1-week delivery iteration × 5 (4 delivery + 1 IP).
-  # Pokud preferujete classic SAFe (2-week iter × 5 = 10 weeks), nastavte
-  # iteration_weeks: 2 a pi_weeks: 10.
-  iteration_weeks: 1
-  pi_weeks: 5
-  delivery_iterations_per_pi: 4
-  ip_iterations_per_pi: 1
+#
+# Kadenci (iteration_weeks / pi_iterations) nastav per-PI v `pi:` bloku
+# .edpa/iterations/PI-*.yaml, ne tady. Viz docs/cadence.md.
 
 teams:
   - id: "kashealth"

diff --git a/docs/playbook.md b/docs/playbook.md
@@ -108,7 +108,7 @@ my-project/
   .edpa/
     config/
       edpa.yaml          # Projekt + governance (zdroj verze metodiky)
-      people.yaml        # Tym a kapacity (cadence, teams, people)
+      people.yaml        # Tym a kapacity (teams, people)
       cw_heuristics.yaml # CW signalove vahy
       id_counters.yaml   # Citace ID (S-1, E-1, F-1, ...)
     backlog/
@@ -152,19 +152,9 @@ my-project/
 cp .edpa/engine/templates/people.yaml.tmpl .edpa/config/people.yaml
 ```
 
-Upravit `.edpa/config/people.yaml`. Tady zije **cadence + teams + people** -- zadny separatni registr, vse je v jednom souboru:
+Upravit `.edpa/config/people.yaml`. Tady ziji **teams + people** -- zadny separatni registr. Kadence (iteration_weeks / pi_iterations) se nastavuje per-PI v `pi:` bloku `.edpa/iterations/PI-*.yaml`, pripadne se odvodi z `weeks:`/dat iteraci -- viz [cadence.md](cadence.md):
 
 ```yaml
-cadence:
-  # AI-native default: 5-tydenni PI = 4 dodavkove iterace po 1 tydnu + 1 IP
-  # tyden na dluh, prioritizaci a PI planning (s AI zvladnutelne za den).
-  # Classic SAFe (2-tydenni iterace, 10-tydenni PI) zustava podporovany —
-  # nastav iteration_weeks: 2 + pi_weeks: 10.
-  iteration_weeks: 1                    # 1 (AI-native, default) nebo 2 (classic)
-  pi_weeks: 5                           # 5 (AI-native, default) nebo 10 (classic)
-  delivery_iterations_per_pi: 4         # PI minus IP iterace
-  ip_iterations_per_pi: 1              # Innovation & Planning
-
 teams:
   - id: "Muj Tym"
     planning_factor: 0.8               # Planujeme na 80% kapacity
@@ -778,7 +768,7 @@ V2 ma **jediny** volitelny workflow (jen s `--with-ci`):
 ### Den 1
 
 - [ ] Engine vendorovany do `.edpa/engine/` (`/edpa:setup` nebo `project_setup.py`)
-- [ ] `.edpa/config/people.yaml` -- tym s rolemi, FTE, kapacitami (cadence + teams + people)
+- [ ] `.edpa/config/people.yaml` -- tym s rolemi, FTE, kapacitami (teams + people)
 - [ ] `.edpa/config/edpa.yaml` -- nazev projektu, governance, naming
 - [ ] `.edpa/config/cw_heuristics.yaml` -- vychozi signalove vahy (ze sablony)
 - [ ] `.edpa/config/id_counters.yaml` -- naseto

diff --git a/docs/quick-start.md b/docs/quick-start.md
@@ -31,10 +31,6 @@ cp plugin/edpa/templates/people.yaml.tmpl .edpa/config/people.yaml
 
 Edit `.edpa/config/people.yaml` with your team:
 ```yaml
-cadence:
-  iteration_weeks: 1        # 1-week iterations (AI-native default; use 2 for classic SAFe)
-  pi_weeks: 5               # 5-week planning intervals (4 delivery + 1 IP)
-
 people:
   - id: alice
     name: "Alice Smith"

diff --git a/plugin/.claude-plugin/plugin.json b/plugin/.claude-plugin/plugin.json
@@ -1,6 +1,6 @@
 {
   "name": "edpa",
-  "version": "2.1.9",
+  "version": "2.1.10",
   "description": "EDPA — Evidence-Driven Proportional Allocation. Derive hours from local git evidence (commits, yaml edits, status transitions). Zero timesheets, mathematical guarantee, Monte Carlo calibrated CW weights. Local-first: .edpa/backlog/ YAML as source of truth, git as the audit trail. GitHub Projects sync optional.",
   "author": {
     "name": "TECHNOMATON",

diff --git a/plugin/edpa/scripts/backlog.py b/plugin/edpa/scripts/backlog.py
@@ -183,7 +183,8 @@ def load_backlog(root):
     """
     edpa = root / ".edpa"
 
-    # Load team metadata (people, teams, cadence)
+    # Load team metadata (people, teams). The cadence: block is not read here —
+    # effective cadence is derived from .edpa/iterations/PI-*.yaml (see _pi_loader).
     people_path = edpa / "config" / "people.yaml"
     backlog = yaml.safe_load(open(people_path, encoding="utf-8")) if people_path.exists() else {}
 

diff --git a/plugin/edpa/scripts/mcp_server.py b/plugin/edpa/scripts/mcp_server.py
@@ -1407,7 +1407,7 @@ async def list_resources() -> list[Resource]:
     resources = []
     if edpa_root:
         if (edpa_root / "config" / "edpa.yaml").exists():
-            resources.append(Resource(uri="edpa://config", name="EDPA Configuration", description="Master config: PI, iterations, cadence, sync settings", mimeType="application/x-yaml"))
+            resources.append(Resource(uri="edpa://config", name="EDPA Configuration", description="Project config: name, funding, organizations, governance, naming, issue types", mimeType="application/x-yaml"))
         if (edpa_root / "config" / "people.yaml").exists():
             resources.append(Resource(uri="edpa://people", name="EDPA Team Registry", description="Team members, roles, FTE, capacity", mimeType="application/x-yaml"))
         # Add iteration resources for each iteration

diff --git a/plugin/edpa/templates/edpa.yaml.tmpl b/plugin/edpa/templates/edpa.yaml.tmpl
@@ -64,7 +64,7 @@ project:
 
 governance:
   # Auto-stamped to the installed plugin version by install.sh / project_setup.py.
-  methodology: "EDPA 2.1.9"
+  methodology: "EDPA 2.1.10"
   # v1.14: single calculation path. v1.17: extended with yaml_edit
   # signals — every commit on .edpa/backlog/<typ>/<id>.md in the
   # iteration window contributes structural credit (create / block_add

diff --git a/plugin/edpa/templates/people.yaml.tmpl b/plugin/edpa/templates/people.yaml.tmpl
@@ -2,16 +2,9 @@
 # Edit this file with your team details
 # Updated at each Iteration Planning
 
-cadence:
-  # Default is AI-native: 5-week PI = 4× 1-week delivery iteration + 1× 1-week IP.
-  # The IP iteration absorbs leftover work, debt, prioritization, and PI planning
-  # itself (compressible to a single day with AI-assisted ceremonies).
-  # Set iteration_weeks: 2 + pi_weeks: 10 + delivery_iterations_per_pi: 4 for
-  # classic SAFe cadence on slower-moving teams.
-  iteration_weeks: 1                          # 1 (AI-native, default) or 2 (classic SAFe)
-  pi_weeks: 5                                 # 5 (AI-native, default) or 10 (classic)
-  delivery_iterations_per_pi: 4               # PI minus IP iteration
-  ip_iterations_per_pi: 1                     # Innovation & Planning
+# Cadence (iteration_weeks / pi_iterations) is NOT configured here. It is set
+# per Planning Interval in the `pi:` block of .edpa/iterations/PI-*.yaml, or
+# auto-derived from each iteration's `weeks:`/dates. See docs/cadence.md.
 
 # Teams — planning_factor is a team-level decision, not a cadence property
 teams:
@@ -88,7 +81,7 @@ people:
 # Calculated at runtime by EDPA engine:
 # total_fte: sum of people[].fte
 # total_capacity_per_iteration: sum of people[].capacity_per_iteration
-# total_capacity_per_pi: total_capacity_per_iteration × delivery_iterations_per_pi
+# total_capacity_per_pi: total_capacity_per_iteration × delivery iterations in the PI
 
 # ─────────────────────────────────────────────────────────────────────
 # Backlog item evidence schema (v1.11 — referenced from .edpa/backlog/**/*.yaml)

diff --git a/tests/e2e_v2_full/fixtures/edpa.yaml b/tests/e2e_v2_full/fixtures/edpa.yaml
@@ -58,15 +58,6 @@ naming:
     defect: "D"
     event: "EV"
 
-# AI-native cadence: 1-week iterations, 5-week PI (4 delivery + 1 IP).
-# Must mirror the cadence block in people.yaml — the test harness
-# verifies this match as part of Wave C invariant checks.
-cadence:
-  iteration_weeks: 1
-  pi_weeks: 5
-  delivery_iterations_per_pi: 4
-  ip_iterations_per_pi: 1
-
 # Native GitHub Issue Types (org-level, not labels)
 issue_types:
   Initiative: "Business case, investment proposal"

diff --git a/tests/e2e_v2_full/fixtures/people.yaml b/tests/e2e_v2_full/fixtures/people.yaml
@@ -5,13 +5,6 @@
 # Total capacity per iteration must sum to exactly 144 hours, which
 # Wave C asserts as an invariant.
 
-cadence:
-  # Must match the cadence block in edpa.yaml (1-week iters, 5-week PI).
-  iteration_weeks: 1
-  pi_weeks: 5
-  delivery_iterations_per_pi: 4
-  ip_iterations_per_pi: 1
-
 # Single team for the synthetic project; planning_factor 0.8 leaves
 # 20% buffer per the AI-native defaults.
 teams:

diff --git a/tests/test_capacity_overrides.py b/tests/test_capacity_overrides.py
@@ -48,8 +48,6 @@ def _seed_minimal_edpa(tmp_path: Path, *, iteration_id="PI-2026-1.1",
     (edpa / "backlog" / "features").mkdir(parents=True)
 
     (edpa / "config" / "people.yaml").write_text(yaml.safe_dump({
-        "cadence": {"iteration_weeks": 1, "pi_weeks": 5,
-                    "delivery_iterations_per_pi": 4, "ip_iterations_per_pi": 1},
         "teams": [{"id": "T", "planning_factor": 0.8}],
         "people": [
             {"id": "bob-dev",   "name": "Bob",   "role": "Dev",  "team": "T",

diff --git a/web/src/pages/en/guide.astro b/web/src/pages/en/guide.astro
@@ -84,15 +84,15 @@ Next: /edpa setup "Project Name"`,
 /edpa:setup "Project Name" --with-ci`,
     code: `# Manual alternative (outside Claude Code):
 # 1. Edit .edpa/config/people.yaml (team, roles, FTE, capacity)
-# 2. Edit .edpa/config/edpa.yaml (project name, cadence)
+# 2. Edit .edpa/config/edpa.yaml (project name, governance, naming)
 # 3. Seed ID counters:
 python3 .edpa/engine/scripts/setup.py
 
 # Optional — PR-signal CI workflow:
 python3 .edpa/engine/scripts/setup.py --with-ci`,
     result: `✓ Engine vendored: .edpa/engine/{scripts,templates,schemas}
 ✓ Team: 3 members configured (2.8 FTE)
-✓ Config: .edpa/config/edpa.yaml (project + cadence)
+✓ Config: .edpa/config/edpa.yaml (project + governance)
 ✓ Config: .edpa/config/people.yaml (team)
 ✓ Seeded: .edpa/config/id_counters.yaml
 ✓ CW heuristics: .edpa/config/cw_heuristics.yaml

diff --git a/web/src/pages/en/playbook.astro b/web/src/pages/en/playbook.astro
@@ -101,7 +101,7 @@ curl -fsSL https://edpa.technomaton.com/install.sh | sh</code></pre>
   .edpa/
     config/
       edpa.yaml          # Project + governance (source of methodology version)
-      people.yaml        # Team and capacities (cadence, teams, people)
+      people.yaml        # Team and capacities (teams, people)
       cw_heuristics.yaml # CW signal weights
       id_counters.yaml   # ID counters (S-1, E-1, F-1, ...)
     backlog/
@@ -138,18 +138,8 @@ curl -fsSL https://edpa.technomaton.com/install.sh | sh</code></pre>
     <h3>1.2 Configure the team and capacities (people.yaml)</h3>
     <p><code>project_setup.py</code> already seeds <code>.edpa/config/people.yaml</code> from the template. Or manually:</p>
     <pre><code>cp .edpa/engine/templates/people.yaml.tmpl .edpa/config/people.yaml</code></pre>
-    <p>Edit <code>.edpa/config/people.yaml</code>. This is where <strong>cadence + teams + people</strong> live — no separate registry, everything is in one file:</p>
-    <pre><code>cadence:
-  # AI-native default: 5-week PI = 4 delivery iterations of 1 week + 1 IP
-  # (a week for debt, prioritization and PI planning — with AI, doable in a day).
-  # Classic SAFe (2-week iterations, 10-week PI) stays supported —
-  # set iteration_weeks: 2 + pi_weeks: 10.
-  iteration_weeks: 1                    # 1 (AI-native, default) or 2 (classic)
-  pi_weeks: 5                           # 5 (AI-native, default) or 10 (classic)
-  delivery_iterations_per_pi: 4         # PI minus IP iteration
-  ip_iterations_per_pi: 1              # Innovation &amp; Planning
-
-teams:
+    <p>Edit <code>.edpa/config/people.yaml</code>. This is where <strong>teams + people</strong> live — no separate registry. Set cadence (iteration_weeks / pi_iterations) per-PI in the <code>pi:</code> block of <code>.edpa/iterations/PI-*.yaml</code>:</p>
+    <pre><code>teams:
   - id: "My Team"
     planning_factor: 0.8               # Plan to 80% of capacity
 
@@ -624,7 +614,7 @@ python3 .edpa/engine/scripts/reports.py PI-2026-1.1</code></pre>
     <h3>Day 1</h3>
     <ul class="checklist">
       <li>Engine vendored into <code>.edpa/engine/</code> (<code>/edpa:setup</code> or <code>project_setup.py</code>)</li>
-      <li><code>.edpa/config/people.yaml</code> — team with roles, FTE, capacities (cadence + teams + people)</li>
+      <li><code>.edpa/config/people.yaml</code> — team with roles, FTE, capacities (teams + people)</li>
       <li><code>.edpa/config/edpa.yaml</code> — project name, governance, naming</li>
       <li><code>.edpa/config/cw_heuristics.yaml</code> — default signal weights (from the template)</li>
       <li><code>.edpa/config/id_counters.yaml</code> — seeded</li>

diff --git a/web/src/pages/en/setup.astro b/web/src/pages/en/setup.astro
@@ -879,7 +879,6 @@ import Layout from '../../layouts/Layout.astro';
     function generateConfigYaml(): string {
       const org = state.githubOrg || 'my-org';
       const repo = state.githubRepo || 'my-project';
-      const fteCap = getFullCapacity();
       const lines = [
         '# EDPA Configuration',
         `# Generated: ${new Date().toISOString().split('T')[0]}`,
@@ -893,13 +892,6 @@ import Layout from '../../layouts/Layout.astro';
         `  org: "${org}"`,
         `  repo: "${repo}"`,
         '',
-        'cadence:',
-        `  iteration_weeks: ${state.iterWeeks}`,
-        `  delivery_iterations: ${state.deliveryIters}`,
-        `  ip_iterations: ${state.ipIters}`,
-        `  pi_weeks: ${(state.deliveryIters + state.ipIters) * state.iterWeeks}`,
-        `  capacity_per_fte: ${fteCap}`,
-        '',
         'scoring:',
         '  contribution_weights:',
         '    assignee: 1.0',