test(cli): move tests/ out of tools/, separate marmot vs dm

`tools/` is for libraries (arti-build); shell-based interop harnesses
that drive the `amy` binary belong with the CLI module. New layout:

  cli/tests/
  ├── lib.sh                       # shared logging + result tracking
  ├── headless/helpers.sh          # shared amy_a / amy_json / assertions
  ├── marmot/                      # MLS group-messaging interop (vs whitenoise-rs)
  │   ├── marmot-interop.sh
  │   ├── marmot-interop-headless.sh
  │   ├── setup.sh
  │   ├── tests-create.sh
  │   ├── tests-manage.sh
  │   ├── tests-extras.sh
  │   └── patches/
  └── dm/                          # NIP-17 DM interop (amy ↔ amy)
      ├── dm-interop-headless.sh
      ├── setup.sh
      └── tests-dm.sh

Per-suite changes:
- Each suite owns its own setup.sh; previously the Marmot setup was at
  headless/setup.sh and the DM setup at headless/setup-dm.sh, which
  implied shared infra they don't actually share.
- `cli/tests/lib.sh` and `cli/tests/headless/helpers.sh` are the only
  shared bits across suites.
- The DM harness still reuses Marmot's `start_local_relay` /
  `stop_local_relay` (sourced via `../marmot/setup.sh`) — same relay
  binary, no duplication.
- All sourcing paths updated; REPO_ROOT now climbs three levels (was
  two when the harness lived under `tools/`); patches/ is now a sibling
  of marmot/setup.sh rather than under marmot/headless/.
- `.gitignore` updated to cover both suites' state dirs.

cli/ROADMAP.md and cli/tests/README.md updated to point at the new paths.
No behavioural change — every test runs the same code; only the file
layout and source paths moved.

cli/ROADMAP.md

@@ -95,11 +95,11 @@ move anything, re-audit — you're probably duplicating logic.
 8. **Distribution** — Homebrew + Scoop + `.deb` in the same release
    pipeline as desktop. Plan: `cli/plans/2026-04-21-cli-distribution.md`.
 9. **Test suite** — end-to-end against a local relay. Marmot interop is
-   covered by `tools/marmot-interop/marmot-interop-headless.sh`; NIP-17
+   covered by `cli/tests/marmot/marmot-interop-headless.sh`; NIP-17 DM
-   DM interop between two `amy` clients is covered by
+   interop between two `amy` clients is covered by
-   `tools/marmot-interop/dm-interop-headless.sh` (text + file + strict
+   `cli/tests/dm/dm-interop-headless.sh` (text + file + strict 10050 +
-   10050 + fallback + cursor-advance). Neither runs in CI yet (both
+   fallback + cursor-advance). Neither runs in CI yet (both need Rust +
-   need Rust + ~3 min cold relay build).
+   ~3 min cold relay build).
 10. **Everything else in the matrix.**
 
 ---

cli/tests/.gitignore

@@ -0,0 +1,3 @@
+marmot/state/
+marmot/state-headless/
+dm/state-dm-headless/

cli/tests/README.md

@@ -1,22 +1,45 @@
-# Interop Test Harnesses
+# amy CLI test harnesses
 
-Two flavours of the Marmot harness, same scenarios:
+Shell-based end-to-end harnesses that drive the `amy` CLI binary against a
+loopback `nostr-rs-relay`. Layout:
 
-- **`marmot-interop.sh`** — interactive. Drives B/C via `wn` and **prompts the
+```
-  human** to perform each Amethyst-side step in the mobile UI (Identity A).
+cli/tests/
-  Use this for final UI verification.
+├── lib.sh                 # shared logging, results, assertions
-- **`marmot-interop-headless.sh`** — zero prompts. Drives A via the `amy` CLI
+├── headless/              # shared bits used by every harness
-  (`./gradlew :cli:installDist`) and B/C via `wn`. Runs every scenario
+│   └── helpers.sh
-  end-to-end and exits with a pass/fail summary. Use this for CI and for
+├── marmot/                # Marmot / MLS group-messaging interop
-  iterating on the Nostr/Marmot plumbing without needing to touch a phone.
+│   ├── marmot-interop.sh           # interactive — prompts Amethyst Android UI
+│   ├── marmot-interop-headless.sh  # zero-prompt
+│   ├── setup.sh                    # preflight + wn + relay + identities
+│   ├── tests-create.sh             # tests 01–05
+│   ├── tests-manage.sh             # tests 06–08, 11
+│   ├── tests-extras.sh             # tests 09, 10, 12, 13
+│   └── patches/                    # whitenoise-rs harness patches
+└── dm/                    # NIP-17 DM interop (amy ↔ amy)
+    ├── dm-interop-headless.sh
+    ├── setup.sh                    # preflight + identities
+    └── tests-dm.sh
+```
+
+The Marmot harnesses come in two flavours, same scenarios:
+
+- **`marmot/marmot-interop.sh`** — interactive. Drives B/C via `wn` and
+  **prompts the human** to perform each Amethyst-side step in the mobile UI
+  (Identity A). Use this for final UI verification.
+- **`marmot/marmot-interop-headless.sh`** — zero prompts. Drives A via the
+  `amy` CLI (`./gradlew :cli:installDist`) and B/C via `wn`. Runs every
+  scenario end-to-end and exits with a pass/fail summary. Use this for CI
+  and for iterating on the Nostr/Marmot plumbing without needing to touch a
+  phone.
 
 A third, slimmer harness covers the NIP-17 DM surface:
 
-- **`dm-interop-headless.sh`** — two `amy` processes (Identity A and
+- **`dm/dm-interop-headless.sh`** — two `amy` processes (Identity A and
   Identity D) exchange NIP-17 DMs through the loopback nostr-rs-relay.
   No whitenoise-rs required — only `amy` and the relay binary (which
   is shared with the Marmot harness's checkout at
-  `state-headless/nostr-rs-relay/`).
+  `marmot/state-headless/nostr-rs-relay/`).
 
 Both Marmot harnesses validate Amethyst against **whitenoise-rs**
 (https://github.com/marmot-protocol/whitenoise-rs), the reference Rust
@@ -42,7 +65,7 @@ the end of the run.
 | 12 | Offline catch-up / replay | – |
 | 13 | KeyPackage rotation | – |
 
-### DM (amy ↔ amy, NIP-17) — `dm-interop-headless.sh`
+### DM (amy ↔ amy, NIP-17) — `dm/dm-interop-headless.sh`
 
 | # | Test |
 |---|---|

cli/tests/dm/dm-interop-headless.sh

@@ -11,7 +11,8 @@
 set -uo pipefail
 
 SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)"
-REPO_ROOT="$(cd -- "$SCRIPT_DIR/../.." && pwd)"
+REPO_ROOT="$(cd -- "$SCRIPT_DIR/../../.." && pwd)"
+TESTS_DIR="$(cd -- "$SCRIPT_DIR/.." && pwd)"
 STATE_DIR="$SCRIPT_DIR/state-dm-headless"
 LOG_DIR="$STATE_DIR/logs"
 A_DIR="$STATE_DIR/A"
@@ -26,7 +27,7 @@ AMY_BIN="$REPO_ROOT/cli/build/install/amy/bin/amy"
 # Share the nostr-rs-relay checkout with the Marmot harness to avoid
 # rebuilding it twice. Override RELAY_REPO / RELAY_DATA if you want full
 # isolation between runs.
-RELAY_REPO="${RELAY_REPO:-$SCRIPT_DIR/state-headless/nostr-rs-relay}"
+RELAY_REPO="${RELAY_REPO:-$TESTS_DIR/marmot/state-headless/nostr-rs-relay}"
 RELAY_BIN="$RELAY_REPO/target/release/nostr-rs-relay"
 RELAY_DATA="$STATE_DIR/relay"
 RELAY_PORT="${RELAY_PORT:-8090}"
@@ -54,22 +55,23 @@ mkdir -p "$STATE_DIR" "$LOG_DIR" "$A_DIR" "$D_DIR"
 : >"$LOG_FILE"
 : >"$RESULTS_FILE"
 
-# Reuse the logging + result helpers from the Marmot harness. lib.sh
+# Reuse the logging + result helpers shared between every harness.
-# also declares wn-specific helpers; they're harmless when unused.
+# lib.sh declares wn-specific helpers too — harmless when unused.
-# shellcheck source=lib.sh
+# shellcheck source=../lib.sh
-source "$SCRIPT_DIR/lib.sh"
+source "$TESTS_DIR/lib.sh"
 
-# Reuse start_local_relay / stop_local_relay from setup.sh. preflight()
+# Reuse start_local_relay / stop_local_relay from the Marmot harness's
-# there also builds whitenoise-rs, which we don't need — setup-dm.sh
+# setup.sh — the relay lifecycle is identical. preflight() there also
+# builds whitenoise-rs, which we don't need; setup.sh in this dir
 # defines a slimmer preflight_dm().
-# shellcheck source=headless/setup.sh
+# shellcheck source=../marmot/setup.sh
-source "$SCRIPT_DIR/headless/setup.sh"
+source "$TESTS_DIR/marmot/setup.sh"
-# shellcheck source=headless/setup-dm.sh
+# shellcheck source=setup.sh
-source "$SCRIPT_DIR/headless/setup-dm.sh"
+source "$SCRIPT_DIR/setup.sh"
-# shellcheck source=headless/helpers.sh
+# shellcheck source=../headless/helpers.sh
-source "$SCRIPT_DIR/headless/helpers.sh"
+source "$TESTS_DIR/headless/helpers.sh"
-# shellcheck source=headless/tests-dm.sh
+# shellcheck source=tests-dm.sh
-source "$SCRIPT_DIR/headless/tests-dm.sh"
+source "$SCRIPT_DIR/tests-dm.sh"
 
 cleanup() {
   local rc=$?

cli/tests/dm/setup.sh

@@ -1,6 +1,6 @@
 # shellcheck shell=bash
 #
-# headless/setup-dm.sh — amy-only preflight + identity bootstrap for the
+# setup.sh — amy-only preflight + identity bootstrap for the
 # NIP-17 DM interop harness. Much slimmer than the Marmot setup:
 #
 #   - Builds `amy` (same retry-on-503 logic as setup.sh).
@@ -11,7 +11,7 @@
 #     recipient-inbox routing has something to resolve to.
 #
 # The heavy `start_local_relay` / `stop_local_relay` helpers live in
-# headless/setup.sh and are sourced by the top-level harness.
+# the Marmot harness's setup.sh and are sourced by the top-level harness.
 
 # --- preflight (amy + relay only, no wn / Marmot patches) -------------------
 preflight_dm() {

cli/tests/dm/tests-dm.sh

@@ -1,6 +1,6 @@
 # shellcheck shell=bash
 #
-# headless/tests-dm.sh — NIP-17 DM interop tests for two `amy` clients.
+# tests-dm.sh — NIP-17 DM interop tests for two `amy` clients.
 #
 # Identity A (sender) and Identity D (recipient) each live in their own
 # --data-dir and share one loopback nostr-rs-relay. Tests cover:

cli/tests/headless/helpers.sh

@@ -1,6 +1,6 @@
 # shellcheck shell=bash
 #
-# headless/helpers.sh — thin wrappers that keep the per-test code tight.
+# helpers.sh — thin wrappers that keep the per-test code tight.
 
 # --- amy wrapper -------------------------------------------------------------
 amy_a() { "$AMY_BIN" --data-dir "$A_DIR" "$@"; }

cli/tests/marmot/marmot-interop-headless.sh

@@ -14,7 +14,8 @@
 set -uo pipefail
 
 SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)"
-REPO_ROOT="$(cd -- "$SCRIPT_DIR/../.." && pwd)"
+REPO_ROOT="$(cd -- "$SCRIPT_DIR/../../.." && pwd)"
+TESTS_DIR="$(cd -- "$SCRIPT_DIR/.." && pwd)"
 STATE_DIR="$SCRIPT_DIR/state-headless"
 LOG_DIR="$STATE_DIR/logs"
 A_DIR="$STATE_DIR/A"
@@ -66,19 +67,19 @@ mkdir -p "$STATE_DIR" "$LOG_DIR" "$A_DIR" "$B_DIR/logs" "$C_DIR/logs"
 : >"$RESULTS_FILE"
 
 # Reuse colours / logging / dump_daemon_diagnostics from the interactive harness.
-# shellcheck source=lib.sh
+# shellcheck source=../lib.sh
-source "$SCRIPT_DIR/lib.sh"
+source "$TESTS_DIR/lib.sh"
 
-# shellcheck source=headless/setup.sh
+# shellcheck source=setup.sh
-source "$SCRIPT_DIR/headless/setup.sh"
+source "$SCRIPT_DIR/setup.sh"
-# shellcheck source=headless/helpers.sh
+# shellcheck source=../headless/helpers.sh
-source "$SCRIPT_DIR/headless/helpers.sh"
+source "$TESTS_DIR/headless/helpers.sh"
-# shellcheck source=headless/tests-create.sh
+# shellcheck source=tests-create.sh
-source "$SCRIPT_DIR/headless/tests-create.sh"
+source "$SCRIPT_DIR/tests-create.sh"
-# shellcheck source=headless/tests-manage.sh
+# shellcheck source=tests-manage.sh
-source "$SCRIPT_DIR/headless/tests-manage.sh"
+source "$SCRIPT_DIR/tests-manage.sh"
-# shellcheck source=headless/tests-extras.sh
+# shellcheck source=tests-extras.sh
-source "$SCRIPT_DIR/headless/tests-extras.sh"
+source "$SCRIPT_DIR/tests-extras.sh"
 
 # Make sure Ctrl+C / SIGTERM / SIGHUP all run the full cleanup path —
 # otherwise wnd is nohup'd and keeps running after the script dies,

cli/tests/marmot/marmot-interop.sh

@@ -10,6 +10,7 @@
 set -uo pipefail
 
 SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)"
+TESTS_DIR="$(cd -- "$SCRIPT_DIR/.." && pwd)"
 STATE_DIR="$SCRIPT_DIR/state"
 LOG_DIR="$STATE_DIR/logs"
 B_DIR="$STATE_DIR/B"
@@ -75,8 +76,8 @@ mkdir -p "$STATE_DIR" "$LOG_DIR" "$B_DIR/logs" "$C_DIR/logs"
 : >"$LOG_FILE"
 : >"$RESULTS_FILE"
 
-# shellcheck source=lib.sh
+# shellcheck source=../lib.sh
-source "$SCRIPT_DIR/lib.sh"
+source "$TESTS_DIR/lib.sh"
 
 # --- preflight ---------------------------------------------------------------
 preflight() {

cli/tests/marmot/setup.sh

@@ -1,6 +1,6 @@
 # shellcheck shell=bash
 #
-# headless/setup.sh — preflight + daemon lifecycle + identity bootstrap.
+# setup.sh — preflight + daemon lifecycle + identity bootstrap.
 # Sourced from marmot-interop-headless.sh.
 
 # --- preflight ---------------------------------------------------------------
@@ -90,7 +90,7 @@ preflight() {
     if [[ ! -f "$marker" ]]; then
       step "patching whitenoise-rs: $name"
       if ( cd "$WN_REPO" && patch -p1 --forward --reject-file=- \
-             <"$SCRIPT_DIR/headless/patches/$name" >>"$LOG_FILE" 2>&1 ); then
+             <"$SCRIPT_DIR/patches/$name" >>"$LOG_FILE" 2>&1 ); then
         touch "$marker"
         # Invalidate the previous build so the patched source is picked up.
         rm -f "$WN_BIN" "$WND_BIN"

cli/tests/marmot/tests-create.sh

@@ -1,6 +1,6 @@
 # shellcheck shell=bash
 #
-# headless/tests-create.sh — tests 01..05.
+# tests-create.sh — tests 01..05.
 # Focus: KeyPackage discovery, group creation, invites, initial messages.
 
 test_01_keypackage_discovery() {

cli/tests/marmot/tests-extras.sh

@@ -1,6 +1,6 @@
 # shellcheck shell=bash
 #
-# headless/tests-extras.sh — tests 09, 10, 12, 13.
+# tests-extras.sh — tests 09, 10, 12, 13.
 # Focus: reactions/replies, concurrent commits, offline catchup, KP rotation.
 
 test_09_reply_react_unreact() {

cli/tests/marmot/tests-manage.sh

@@ -1,6 +1,6 @@
 # shellcheck shell=bash
 #
-# headless/tests-manage.sh — tests 06, 07, 08, 11.
+# tests-manage.sh — tests 06, 07, 08, 11.
 # Focus: removal, metadata rename, admin promote/demote, leave.
 
 test_06_member_removal() {

tools/marmot-interop/.gitignore

@@ -1,2 +0,0 @@
-state/
-state-headless/