2026-10-28 15:01:55 +00:00
2 changed files with 341 additions and 2 deletions
@@ -81,6 +81,8 @@ Reconnect works the same as the list stream: each SSE message carries an `id:` l
 | POST | `/api/nests/{nest_id}/messages` | `{"author", "text", "to"?, "kind"? (crow header), "refers_to"?, "approval_phrase"? (honored only when the resulting kind is `ask_owner`)}` | 201 with the stored record (server-assigned `kind`, `tagged_by`, `msg_id`). A fresh `ask_owner` reads `resolved: ""`. 404 unknown nest; 400 with `{"message"}` for bad input; 403 with `{"message"}` when the crow author is quarantined; 429 with `{"message","cooldown_s"}` when the poster is rate-limited. |
 | GET | `/api/events` | `?nest_id=<id>` (parallel to `?list_id=`) | SSE stream of `nest.message` events for that nest. |

+Service roles (Elder, Quill, Boss) are addressable in every nest but are never members. A `request` addressed to Boss triggers one `say` back to the requester: the outcome of the order (a new crow joined, or a refusal naming the reason — cap reached, name taken, persona unknown). Boss works through the flock like any crow: a saturated flock means no Boss reply, and no registry write.
+
 ### Tagging [decided 2026-10-27]

 Jev choice-tags every posted message (`kind` question). Above its high band, Jev's tag stands (`tagged_by: "jev"`). If Jev is unsure, the crow's own `kind` header is used (`"header"`). If both are present and disagree, Jev wins and it is recorded as `"disagreement"`. If neither yields a kind, the message is `say` (`"default"`). Every tag is logged with its probabilities so the calibration experiment has data.
@@ -93,6 +95,7 @@ Jev choice-tags every posted message (`kind` question). Above its high band, Jev
 - Watchdog (E3): every crow `claim` or `deliver` that refers to a request gets one Jev call (`Flock.ClassifyWatch`, strategies 4/8 in `JEV_STRATEGIES.md`) asking three noul questions about it against its request. Any score at or above the band, or `answers_request` at or below `1 − band` (band = `CROWS_JEV_THRESHOLD`), is a breach: the Sentry posts a `flag` naming the scores and the crow is quarantined. At most one flag per crow per referent (later breaches are logged, not posted). Owner messages are never inspected.
 - A quarantined crow's POSTs (any crow kind) return 403 with `{"message": "<author> is quarantined pending the owner's decision."}`, and it is skipped by the open-, conspiracy- and directed-response selectors. Quarantine is fail-safe: it sticks until the owner lifts it (`ask_owner` escalation lands with #45).
 - Requests per minute and cooldown reuse the requester strikes in the ledger, keyed by `author`.
+- Spin-ups: at most `MaxSpunCrows` (default 8) registry crows; a spun crow's name may not collide with the persona roster or the registry; starting trust is clamped to `regular` at most.

 ## Settled details

@@ -114,10 +117,15 @@ A relay (D3, task #40): when the owner's request asks another named member crow
 - The default nest is `todo`, with an empty core: its crows come from the whole flock, so nothing there is ever "borrowed".
 - Crows are picked core-first (rotating within the core), then borrowed from outside it (rotating), still capped by the flock guardrails.
 - Nests persist on the server (`.crows/nests.json`).
+- Boss (Chief of Staff crow, CHATROOM_SPEC section 6) is a service role like Elder and Quill: a `request` addressed to it gets one `say` from Boss naming the crow and the outcome. Asked to spin up a crow, Boss answers with a structured order (`SPIN {"name","persona","trust"}`) that the harness executes; everything else gets a plain reply.
+
+### Spun crows
+
+Crows the Boss creates live in a **crow registry** (`.crows/crows.json`), separate from the persona roster. A registry crow is a name, the backing persona whose habits it runs on, and its starting trust. Once created it is an ordinary crow: the same trust ledger and guardrails apply. Spun trust is capped at `regular` — earned trust can't be minted. Registry size is capped by `MaxSpunCrows` (default 8).

 | Method | Path | Body | Returns |
 |---|---|---|---|
-| GET | `/api/nests` | | `{"roster": ["Hugin", ...], "nests": [{"nest_id", "name", "crows": [...]}]}`; roster is the 8 working crows (not Sentry, Elder or Quill) |
+| GET | `/api/nests` | | `{"roster": ["Hugin", ...], "nests": [{"nest_id", "name", "crows": [...]}]}`; roster is the 8 working crows (not Sentry, Elder, Quill or Boss) |
 | POST | `/api/nests` | `{"name": "Launch copy", "crows": ["Pepper", "Inky"]}` | 201 with the nest; 400 with `{"message"}` if invalid |
 | POST | `/api/lists` | adds optional `"nest_id"` (default `todo`) | 404 with `{"message"}` for an unknown nest |

@@ -131,7 +139,8 @@ Event changes:

 | Method | Path | Returns |
 |---|---|---|
-| GET | `/api/crows` | `[{"name", "epithet", "strength", "backstory"}]` for all ten crows |
+| GET | `/api/personas` | `[{"name", "epithet", "strength", "backstory"}]` for every persona crow: the 8 workers plus Sentry, Elder, Quill and Boss |
+| GET | `/api/crows` | `{"crows": [{"name", "persona", "trust"}]}` — the spun-crow registry only (never persona crows), sorted by name. `persona` is the backing persona it runs on; `trust` is the level it started at (`"stranger"` or `"regular"`). An empty registry, or a server with none, returns `{"crows": []}`. |

 | Event | Payload (besides list_id, ts) | Meaning |
 |---|---|---|
@@ -0,0 +1,330 @@
+package main
+
+import (
+	"context"
+	"encoding/json"
+	"net/http"
+	"net/http/httptest"
+	"path/filepath"
+	"strings"
+	"sync"
+	"testing"
+	"time"
+)
+
+// bossBrain is a scripted brain that records how often it was called and hands
+// back one canned reply per call.
+type bossBrain struct {
+	mu    sync.Mutex
+	calls int
+	resp  []string
+}
+
+func (b *bossBrain) Name() string { return "boss-script" }
+func (b *bossBrain) Ask(ctx context.Context, system, prompt string) (string, int, error) {
+	b.mu.Lock()
+	defer b.mu.Unlock()
+	r := "no reply"
+	if b.calls < len(b.resp) {
+		r = b.resp[b.calls]
+	}
+	b.calls++
+	return r, 10, nil
+}
+
+func (b *bossBrain) count() int {
+	b.mu.Lock()
+	defer b.mu.Unlock()
+	return b.calls
+}
+
+// newBossServer builds a whole server by hand (like newDispatchServer) so the
+// Boss tests can attach a CrowStore at a temp path.
+func newBossServer(t *testing.T, replies ...string) (*Server, *bossBrain) {
+	t.Helper()
+	dir := t.TempDir()
+	ledger, err := LoadLedger(filepath.Join(dir, "ledger.json"))
+	if err != nil {
+		t.Fatal(err)
+	}
+	flog, err := OpenFlightLog(filepath.Join(dir, "flight.jsonl"))
+	if err != nil {
+		t.Fatal(err)
+	}
+	t.Cleanup(func() { flog.Close() })
+	nests, err := LoadNests(filepath.Join(dir, "nests.json"))
+	if err != nil {
+		t.Fatal(err)
+	}
+	msgs, err := LoadMessageStore(filepath.Join(dir, "messages"))
+	if err != nil {
+		t.Fatal(err)
+	}
+	crows, err := LoadCrowStore(filepath.Join(dir, "crows.json"))
+	if err != nil {
+		t.Fatal(err)
+	}
+	ctx, cancel := context.WithCancel(context.Background())
+	t.Cleanup(cancel)
+	g := DefaultGuardrails()
+	bb := &bossBrain{resp: replies}
+	appr := NewWebApprover(time.Minute)
+	p := &Pipeline{Flock: NewFlock(bb, g, flog), Ledger: ledger, Log: flog, G: g, Approver: appr}
+	s := NewServer(ctx, p, appr, nests).WithMessages(msgs).WithCrows(crows)
+	return s, bb
+}
+
+func askBoss(t *testing.T, ts *httptest.Server, nestID, text string) (int, map[string]any) {
+	t.Helper()
+	code, out := post(t, ts.URL+"/api/nests/"+nestID+"/messages",
+		`{"author":"owner","kind":"request","to":"Boss","text":`+jsonString(text)+`}`)
+	if code != http.StatusCreated {
+		t.Fatalf("owner POST to Boss: want 201, got %d %v", code, out)
+	}
+	return code, out
+}
+
+func jsonString(s string) string {
+	b, _ := json.Marshal(s)
+	return string(b)
+}
+
+func registryCrows(t *testing.T, ts *httptest.Server) []map[string]string {
+	t.Helper()
+	resp, err := http.Get(ts.URL + "/api/crows")
+	if err != nil {
+		t.Fatal(err)
+	}
+	defer resp.Body.Close()
+	if resp.StatusCode != http.StatusOK {
+		t.Fatalf("GET /api/crows: want 200, got %d", resp.StatusCode)
+	}
+	var out struct {
+		Crows []map[string]string `json:"crows"`
+	}
+	if err := json.NewDecoder(resp.Body).Decode(&out); err != nil {
+		t.Fatal(err)
+	}
+	return out.Crows
+}
+
+func bossSay(t *testing.T, ts *httptest.Server, nestID string) map[string]any {
+	t.Helper()
+	for _, m := range awaitMessages(t, ts, nestID, 2) {
+		if m["author"] == "Boss" {
+			return m
+		}
+	}
+	t.Fatalf("Boss never answered: %v", nestMessages(t, ts, nestID))
+	return nil
+}
+
+func spunByName(crows []map[string]string, name string) map[string]string {
+	for _, c := range crows {
+		if c["name"] == name {
+			return c
+		}
+	}
+	return nil
+}
+
+func TestBossSpinUpCreatesCrow(t *testing.T) {
+	s, bb := newBossServer(t, `SPIN {"name":"Ash","persona":"Soot","trust":"stranger"}`)
+	ts := serve(s)
+	defer ts.Close()
+	id := newNest(t, ts, "Lab", `"Hugin"`)
+
+	askBoss(t, ts, id, "spin up a crow named Ash")
+	say := bossSay(t, ts, id)
+	if say["kind"] != "say" || say["to"] != "owner" {
+		t.Fatalf("Boss must post a say back to the owner, got %v", say)
+	}
+	if !strings.Contains(say["text"].(string), "Ash") {
+		t.Fatalf("Boss say must name the new crow, got %v", say["text"])
+	}
+
+	got := spunByName(registryCrows(t, ts), "Ash")
+	if got == nil {
+		t.Fatalf("registry must contain Ash, got %v", registryCrows(t, ts))
+	}
+	if got["persona"] != "Soot" || got["trust"] != "stranger" {
+		t.Fatalf("want Ash on Soot at stranger, got %v", got)
+	}
+
+	rec := s.p.Ledger.Crows["Ash|other"]
+	if rec == nil || rec.Level != Stranger {
+		t.Fatalf("ledger must carry Ash|other at Stranger, got %+v", rec)
+	}
+	if bb.count() != 1 {
+		t.Fatalf("one spin-up is exactly one model call, got %d", bb.count())
+	}
+}
+
+func TestSpinUpCapRefused(t *testing.T) {
+	s, _ := newBossServer(t,
+		`SPIN {"name":"Ash","persona":"Soot","trust":"stranger"}`,
+		`SPIN {"name":"Beck","persona":"Jet","trust":"stranger"}`)
+	s.p.G.MaxSpunCrows = 1
+	ts := serve(s)
+	defer ts.Close()
+	id := newNest(t, ts, "Lab", `"Hugin"`)
+
+	askBoss(t, ts, id, "spin up a crow named Ash")
+	bossSay(t, ts, id)
+	if n := len(registryCrows(t, ts)); n != 1 {
+		t.Fatalf("first spin-up must land, registry has %d", n)
+	}
+
+	askBoss(t, ts, id, "spin up a crow named Beck")
+	say := bossSay(t, ts, id)
+	text, _ := say["text"].(string)
+	if !strings.Contains(text, "cap") {
+		t.Fatalf("the refusal must name the cap, got %q", text)
+	}
+	if n := len(registryCrows(t, ts)); n != 1 {
+		t.Fatalf("a refused spin-up must not touch the registry, has %d", n)
+	}
+	if _, ok := s.p.Ledger.Crows["Beck|other"]; ok {
+		t.Fatal("a refused spin-up must not touch the ledger")
+	}
+}
+
+func TestSpinUpDuplicateNameRefused(t *testing.T) {
+	s, _ := newBossServer(t, `SPIN {"name":"Hugin","persona":"Soot","trust":"stranger"}`)
+	ts := serve(s)
+	defer ts.Close()
+	id := newNest(t, ts, "Lab", `"Hugin"`)
+
+	askBoss(t, ts, id, "spin up a crow named Hugin")
+	say := bossSay(t, ts, id)
+	text, _ := say["text"].(string)
+	if !strings.Contains(text, "taken") {
+		t.Fatalf("the refusal must name the taken name, got %q", text)
+	}
+	if n := len(registryCrows(t, ts)); n != 0 {
+		t.Fatalf("registry must stay empty, has %d", n)
+	}
+}
+
+func TestSpinUpTrustGated(t *testing.T) {
+	s, _ := newBossServer(t, `SPIN {"name":"Ash","persona":"Soot","trust":"trusted"}`)
+	ts := serve(s)
+	defer ts.Close()
+	id := newNest(t, ts, "Lab", `"Hugin"`)
+
+	askBoss(t, ts, id, "spin up a crow named Ash, make it trusted")
+	bossSay(t, ts, id)
+
+	got := spunByName(registryCrows(t, ts), "Ash")
+	if got == nil {
+		t.Fatal("Ash must be in the registry")
+	}
+	if got["trust"] != "regular" {
+		t.Fatalf("trusted must clamp to regular, got %q", got["trust"])
+	}
+	rec := s.p.Ledger.Crows["Ash|other"]
+	if rec == nil || rec.Level > Regular {
+		t.Fatalf("ledger level must be at most Regular, got %+v", rec)
+	}
+}
+
+func TestSpinUpGoesThroughFlock(t *testing.T) {
+	s, _ := newBossServer(t, `SPIN {"name":"Ash","persona":"Soot","trust":"stranger"}`)
+	s.p.Flock.sem = make(chan struct{}, 0) // saturated flock: ask must wait, never act
+	ts := serve(s)
+	defer ts.Close()
+	id := newNest(t, ts, "Lab", `"Hugin"`)
+
+	askBoss(t, ts, id, "spin up a crow named Ash")
+	time.Sleep(300 * time.Millisecond)
+
+	for _, m := range nestMessages(t, ts, id) {
+		if m["author"] == "Boss" {
+			t.Fatalf("a saturated flock must not produce a Boss message, got %v", m)
+		}
+	}
+	if n := len(registryCrows(t, ts)); n != 0 {
+		t.Fatalf("a saturated flock must not write the registry, has %d", n)
+	}
+}
+
+func TestBossPlainReply(t *testing.T) {
+	s, _ := newBossServer(t, "The schedule stands. Nothing needs whip.")
+	ts := serve(s)
+	defer ts.Close()
+	id := newNest(t, ts, "Lab", `"Hugin"`)
+
+	askBoss(t, ts, id, "how goes the books")
+	say := bossSay(t, ts, id)
+	if say["kind"] != "say" || say["text"] != "The schedule stands. Nothing needs whip." {
+		t.Fatalf("Boss posts its plain reply as a say, got %v", say)
+	}
+	if n := len(registryCrows(t, ts)); n != 0 {
+		t.Fatalf("a plain reply must not touch the registry, has %d", n)
+	}
+}
+
+func TestSpinUpGarbledRefused(t *testing.T) {
+	s, _ := newBossServer(t, `SPIN {"name": broken json`)
+	ts := serve(s)
+	defer ts.Close()
+	id := newNest(t, ts, "Lab", `"Hugin"`)
+
+	askBoss(t, ts, id, "spin up a crow")
+	say := bossSay(t, ts, id)
+	text, _ := say["text"].(string)
+	if text == "" {
+		t.Fatal("an unreadable order must still get a refusal, got %v", say)
+	}
+	if n := len(registryCrows(t, ts)); n != 0 {
+		t.Fatalf("an unreadable order must not touch the registry, has %d", n)
+	}
+}
+
+func TestPersonasEndpointIncludesBoss(t *testing.T) {
+	ts, _ := newTestNestServer(t)
+	defer ts.Close()
+	resp, err := http.Get(ts.URL + "/api/personas")
+	if err != nil {
+		t.Fatal(err)
+	}
+	defer resp.Body.Close()
+	if resp.StatusCode != http.StatusOK {
+		t.Fatalf("GET /api/personas: want 200, got %d", resp.StatusCode)
+	}
+	var out []map[string]string
+	if err := json.NewDecoder(resp.Body).Decode(&out); err != nil {
+		t.Fatal(err)
+	}
+	names := map[string]bool{}
+	for _, p := range out {
+		names[p["name"]] = true
+	}
+	for _, want := range []string{"Boss", "Quill", "Hugin"} {
+		if !names[want] {
+			t.Fatalf("/api/personas must list every persona crow (missing %s): %v", want, names)
+		}
+	}
+}
+
+func TestCrowsRegistryEmptyWithoutStore(t *testing.T) {
+	ts, _ := newTestNestServer(t) // no WithCrows: the registry reads as empty
+	defer ts.Close()
+	resp, err := http.Get(ts.URL + "/api/crows")
+	if err != nil {
+		t.Fatal(err)
+	}
+	defer resp.Body.Close()
+	if resp.StatusCode != http.StatusOK {
+		t.Fatalf("want 200 with an empty list, got %d", resp.StatusCode)
+	}
+	var out struct {
+		Crows []map[string]string `json:"crows"`
+	}
+	if err := json.NewDecoder(resp.Body).Decode(&out); err != nil {
+		t.Fatal(err)
+	}
+	if len(out.Crows) != 0 {
+		t.Fatalf("want no crows, got %v", out.Crows)
+	}
+}