added /models cmd

2026-08-28 09:30:48 +03:00
parent 7b26d8f485
commit 699bb429fb
3 changed files with 132 additions and 2 deletions
@@ -66,6 +66,7 @@ All implementations read the same `model.cfg` and `system.txt` from the current
   - `/load <id>` — load a saved session (exact id or unique prefix) and continue from there
   - `/compact` — compact context down to the system message and a concise summary using the LLM; the compaction prompt is appended to the conversation to derive the summary, then the conversation is reset to `[system, summary-user-message]` (a fresh prefix, so downstream prompt-cache hits depend on the provider and are not guaranteed)
   - `/cfg <param> [val]` — inspect or update a configuration parameter in `model.cfg` live
+   - `/models` — query the `/models` path on the current inference endpoint and print a plain list of supported model IDs, marking the currently configured model with a leading `* ` (Go port)
   - `!<cmd>` — execute a shell command directly through `shell_exec` without adding the result to the conversation context (Go port)
   - `/help` — show all supported commands
   - `/clear` — reset the conversation to just the system prompt
@@ -112,7 +113,7 @@ Using these rules, everyone can build their own copy of Bantam from scratch in l
 2. Read model parameters from `model.cfg` (`key=value` format) and discover context window size.
 3. Prepare a new message list with the system prompt (`role: "system"`).
 4. Read the first command-line parameter. If non-empty, read user prompt from the specified file. If prefixed with `!`, execute the shell command directly via `shell_exec` and exit. Otherwise, append to `messages` (`role: "user"`), run `AL(cfg, messages)`, display token usage, and exit.
-5. Read user prompt from standard input (with `readline` line editing and history in `~/.bantam_history`; **Ctrl+J** inserts a real newline into the line being edited). If equal to `/quit` or EOF, exit. If equal to `/clear`, reset `messages` to step 3 and return to step 5. If equal to `/save`, write the whole `messages` array to `~/.bantam/sessions/<id>.json` (with an auto-generated summary) and return to step 5. If equal to `/list`, print saved sessions and their summaries and return to step 5. If starting with `/load`, replace `messages` with the saved session's messages (by exact id or unique prefix) and return to step 5. If equal to `/compact`, ask the LLM to summarize the conversation by appending the compaction prompt to derive the summary, replace `messages` with `[system, summary-user-message]`, and return to step 5. If starting with `/cfg`, display the current value (`/cfg <param>`) or update `model.cfg` live (`/cfg <param> <val>`) and return to step 5. If starting with `!`, execute the command directly via `shell_exec` without adding the result to `messages` and return to step 5. If equal to `/help`, print the command list and return to step 5. After every user turn and on exit, auto-save `messages` to `~/.bantam/sessions/autosave.json`.
+5. Read user prompt from standard input (with `readline` line editing and history in `~/.bantam_history`; **Ctrl+J** inserts a real newline into the line being edited). If equal to `/quit` or EOF, exit. If equal to `/clear`, reset `messages` to step 3 and return to step 5. If equal to `/save`, write the whole `messages` array to `~/.bantam/sessions/<id>.json` (with an auto-generated summary) and return to step 5. If equal to `/list`, print saved sessions and their summaries and return to step 5. If starting with `/load`, replace `messages` with the saved session's messages (by exact id or unique prefix) and return to step 5. If equal to `/compact`, ask the LLM to summarize the conversation by appending the compaction prompt to derive the summary, replace `messages` with `[system, summary-user-message]`, and return to step 5. If starting with `/cfg`, display the current value (`/cfg <param>`) or update `model.cfg` live (`/cfg <param> <val>`) and return to step 5. If equal to `/models`, query the `/models` path on the current inference endpoint and print a plain list of supported model IDs (the currently configured model marked with a leading `* `), then return to step 5. If starting with `!`, execute the command directly via `shell_exec` without adding the result to `messages` and return to step 5. If equal to `/help`, print the command list and return to step 5. After every user turn and on exit, auto-save `messages` to `~/.bantam/sessions/autosave.json`.
 6. Append user prompt to `messages` (`role: "user"`), run `AL(cfg, messages)`, display token usage, check 60% context threshold for auto-compaction, and go to step 5.

 ### Agentic loop (`AL(cfg, messages)`) function
@@ -136,6 +136,49 @@ func fetchContextWindow(cfg *Cfg) int {
  return 262144
 }

+// listModels queries the /models path on the configured inference endpoint and
+// returns a plain newline-separated list of supported model IDs. The currently
+// configured model is marked with a leading "* ". On transport or HTTP errors a
+// non-nil error is returned so the caller can surface it.
+func listModels(cfg *Cfg) (string, error) {
+  t := cfg.Timeout
+  if t < 10 { t = 10 }
+  client := &http.Client{Timeout: time.Duration(t) * time.Second}
+  req, err := http.NewRequest("GET", strings.TrimRight(cfg.Endpoint, "/")+"/models", nil)
+  if err != nil { return "", err }
+  req.Header.Set("User-Agent", "Mozilla/5.0 (compatible; Bantam/1.0)")
+  if cfg.APIKey != "" && cfg.APIKey != "-" { req.Header.Set("Authorization", "Bearer "+cfg.APIKey) }
+  resp, err := client.Do(req)
+  if err != nil { return "", err }
+  defer resp.Body.Close()
+  if resp.StatusCode >= 400 {
+    return "", fmt.Errorf("HTTP %d from /models", resp.StatusCode)
+  }
+  var res struct {
+    Data   []map[string]any `json:"data"`
+    Models []map[string]any `json:"models"`
+  }
+  if err := json.NewDecoder(resp.Body).Decode(&res); err != nil {
+    return "", fmt.Errorf("failed to parse /models response: %v", err)
+  }
+  list := res.Data
+  if len(list) == 0 { list = res.Models }
+  if len(list) == 0 {
+    return "No models returned by the endpoint.", nil
+  }
+  var b strings.Builder
+  for _, item := range list {
+    id, _ := item["id"].(string)
+    if id == "" { continue }
+    if id == cfg.Model || strings.EqualFold(id, cfg.Model) {
+      b.WriteString("* " + id + "\n")
+    } else {
+      b.WriteString("  " + id + "\n")
+    }
+  }
+  return b.String(), nil
+}
+
 func getCfg(path string) Cfg {
  cfg := defCfg
  cfg.Raw = map[string]string{
@@ -1042,6 +1085,14 @@ func last(msgs []Message) string {
  return ""
 }

+// lastRole returns the role of the last message in the conversation, or "" if
+// it is empty. It is used to detect when a turn ended on a tool result rather
+// than a normal assistant reply, so the conversation can be auto-continued.
+func lastRole(msgs []Message) string {
+  if len(msgs) == 0 { return "" }
+  return msgs[len(msgs)-1].Role
+}
+
 func AL(ctx context.Context, cfg *Cfg, msgs []Message, sp string, depth int) ([]Message, Usage, error) {
  done := false
  var turnUsage Usage
@@ -1545,6 +1596,11 @@ func main() {
    sigCtx, cancel := signal.NotifyContext(context.Background(), os.Interrupt)
    var usg Usage
    msgs, usg, err = AL(sigCtx, &cfg, msgs, sp, 0)
+    if lastRole(msgs) == "tool" {
+      msgs = append(msgs, Message{Role: "user", Content: strp("continue")})
+      fmt.Println(c("[auto continue: last message was a tool result]", 33))
+      msgs, usg, err = AL(sigCtx, &cfg, msgs, sp, 0)
+    }
    interrupted := sigCtx.Err() != nil
    cancel()
    if err != nil {
@@ -1638,12 +1694,21 @@ func main() {
        fmt.Println(c("Usage: /cfg <param> [val]", 31))
      }
      continue
+    case u == "/models":
+      ml, err := listModels(&cfg)
+      if err != nil {
+        fmt.Println(c("[error querying /models: "+err.Error()+"]", 31))
+        continue
+      }
+      fmt.Println(c("Supported models at "+cfg.Endpoint+":", 1, 36))
+      fmt.Print(renderMD(ml))
+      continue
    case strings.HasPrefix(u, "!"):
      runDirectShell(strings.TrimPrefix(u, "!"), cfg.ShellTimeout)
      continue
    case u == "/help":
      fmt.Println(c("Bantam commands:", 1, 36))
-      for _, kv := range [][2]string{{"/quit", "exit"}, {"/clear", "reset to system prompt"}, {"/save", "save session"}, {"/list", "list sessions"}, {"/load <id>", "load session"}, {"/compact", "compact context"}, {"/cfg <k> [v]", "get/set config"}, {"!<cmd>", "run shell command directly"}, {"/help", "show help"}} {
+      for _, kv := range [][2]string{{"/quit", "exit"}, {"/clear", "reset to system prompt"}, {"/save", "save session"}, {"/list", "list sessions"}, {"/load <id>", "load session"}, {"/compact", "compact context"}, {"/cfg <k> [v]", "get/set config"}, {"!<cmd>", "run shell command directly"}, {"/models", "list models at endpoint"}, {"/help", "show help"}} {
        fmt.Println(c(fmt.Sprintf("  %-15s", kv[0]), 1, 32) + kv[1])
      }
      continue
@@ -1652,6 +1717,11 @@ func main() {
    turnMsgs = append(turnMsgs, Message{Role: "user", Content: strp(u)})
    sigCtx, cancel := signal.NotifyContext(context.Background(), os.Interrupt)
    resMsgs, usg, err := AL(sigCtx, &cfg, turnMsgs, sp, 0)
+    if lastRole(resMsgs) == "tool" {
+      resMsgs = append(resMsgs, Message{Role: "user", Content: strp("continue")})
+      fmt.Println(c("[auto continue: last message was a tool result]", 33))
+      resMsgs, usg, err = AL(sigCtx, &cfg, resMsgs, sp, 0)
+    }
    interrupted := sigCtx.Err() != nil
    cancel()
    if err != nil {
@@ -1845,6 +1845,65 @@ func TestFormatUsage(t *testing.T) {
  }
 }

+func TestListModels(t *testing.T) {
+  srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+    if r.URL.Path != "/models" {
+      http.NotFound(w, r)
+      return
+    }
+    if r.Header.Get("Authorization") != "Bearer secret" {
+      w.WriteHeader(401)
+      return
+    }
+    w.Write([]byte(`{
+      "data": [
+        {"id": "alpha"},
+        {"id": "my-target-model"},
+        {"id": "beta"}
+      ]
+    }`))
+  }))
+  defer srv.Close()
+
+  cfg := defCfg
+  cfg.Endpoint = srv.URL
+  cfg.Model = "my-target-model"
+  cfg.APIKey = "secret"
+
+  out, err := listModels(&cfg)
+  if err != nil {
+    t.Fatalf("unexpected error: %v", err)
+  }
+  lines := strings.Split(strings.TrimRight(out, "\n"), "\n")
+  var got []string
+  for _, l := range lines {
+    if l == "" { continue }
+    got = append(got, l)
+  }
+  want := []string{"  alpha", "* my-target-model", "  beta"}
+  if len(got) != len(want) {
+    t.Fatalf("expected %d lines, got %d: %q", len(want), len(got), out)
+  }
+  for i := range want {
+    if got[i] != want[i] {
+      t.Errorf("line %d: expected %q, got %q", i, want[i], got[i])
+    }
+  }
+}
+
+func TestListModelsHTTPError(t *testing.T) {
+  srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+    w.WriteHeader(500)
+  }))
+  defer srv.Close()
+  cfg := defCfg
+  cfg.Endpoint = srv.URL
+  cfg.APIKey = "-"
+  if _, err := listModels(&cfg); err == nil {
+    t.Fatalf("expected error on HTTP 500, got nil")
+  }
+}
+
 func TestFetchContextWindowFromModelsAPI(t *testing.T) {
  srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
    if r.URL.Path != "/models" {