mirror of
https://github.com/R0m1k3/Loki.git
synced 2026-10-11 17:26:57 +02:00
- mode agent : les outils write/edit/bash résolvent les chemins relatifs dans JEAN_HOME/workspace et le shell y démarre, au lieu du répertoire courant du processus (Bureau, ProgramData\jean\bin). Les chemins absolus demandés explicitement restent honorés. - Windows : le double-clic n'installe plus en silence. Boîte de dialogue unique, raccourcis menu Démarrer + Bureau, puis relance depuis la copie installée — un seul binaire fait foi. Désinstallation : retrait des raccourcis. - update : contrôle des droits AVANT le téléchargement (l'échec tombait sur le fichier temporaire, avec un message brut), message donnant la commande exacte, timeout UI porté de 30 s à 10 min. - jean where + bouton « Où sont mes fichiers ? » : emplacements, et alerte quand le binaire lancé n'est pas celui installé. - icônes : source unique (sys_brand_icon.go) pour la zone de notification, la barre de menus macOS et l'icône du .exe, alignée sur le favicon noir. macOS en icône template (une icône noire figée disparaît sur une barre sombre). - CI : le bundle macOS part de icon.png au lieu de convertir le .ico via sips, dont l'échec était masqué par « || true » ; notes de release versionnées.
269 lines
11 KiB
Go
269 lines
11 KiB
Go
package jean
|
|
|
|
import (
|
|
"context"
|
|
"fmt"
|
|
"os"
|
|
"os/exec"
|
|
"os/user"
|
|
"path/filepath"
|
|
"runtime"
|
|
"strconv"
|
|
"strings"
|
|
"time"
|
|
)
|
|
|
|
const (
|
|
toolDefaultTimeout = 30
|
|
toolMaxTimeout = 300
|
|
toolMaxOutput = 8000 // characters of stdout/stderr returned to the model
|
|
)
|
|
|
|
// baseSystemPrompt is the always-on system preamble. Structured like pi's
|
|
// proven prompt (identity → method → guidelines → concision → date): a concrete,
|
|
// procedural prompt gives the model rails so it stops deliberating forever
|
|
// ("Wait, let me check… Wait, I'll just run it…") and commits to an action.
|
|
// The per-tool "Outil disponible" sections live in machine/skills prompts so
|
|
// they only appear when the matching feature is on.
|
|
func baseSystemPrompt(caps Caps) string {
|
|
hasMem := caps.Mem != MemOff
|
|
// No tool access at all → no agentic preamble. A plain chat model told to
|
|
// "call tools immediately" hallucinates textual tool calls (e.g.
|
|
// default_api:bash) that leak into the answer. Let the user's own system
|
|
// prompt stand alone. (Internet requiert l'agent, donc pas testé ici.)
|
|
if !caps.Agent && !hasMem {
|
|
return ""
|
|
}
|
|
var b strings.Builder
|
|
// Prompt VOLONTAIREMENT court. Un préambule verbeux (longue liste de
|
|
// « guidelines », surtout des méta-instructions sur la réflexion) fait
|
|
// sur-raisonner les modèles à reasoning (Qwen3) : ils émettent leur <think>
|
|
// puis le token de fin SANS appeler d'outil (~25-45 % de tours « morts »
|
|
// mesurés). Une version courte et directe ramène ça à 0 %. NE PAS regonfler.
|
|
// « Jean » avec une majuscule : c'est un nom propre, et le modèle recopie
|
|
// littéralement la casse d'ici quand il se présente (« je suis jean »).
|
|
b.WriteString("You are Jean, an expert assistant operating directly on this machine with real tools.")
|
|
if caps.Mem == MemAlways {
|
|
b.WriteString(" You evolve with every conversation: you actively maintain a persistent memory so nothing useful is lost between sessions.")
|
|
}
|
|
b.WriteString("\n\nTools:\n")
|
|
if caps.Agent {
|
|
b.WriteString("- bash — run a shell command on this machine (inspect files, processes, logs, run scripts). The shell is " + shellName() + ": use its syntax.\n")
|
|
b.WriteString("- write — create or rewrite a file with exact content.\n")
|
|
b.WriteString("- edit — patch a file by exact replacement (old → new, old must be unique).\n")
|
|
}
|
|
if hasMem {
|
|
b.WriteString("- mem_search / mem_read / mem_add / mem_edit — your persistent Markdown memory under MEMORY/.\n")
|
|
}
|
|
// Politique d'usage de la mémoire selon le mode.
|
|
switch caps.Mem {
|
|
case MemAlways:
|
|
b.WriteString("\nManaging your memory is part of the job, not optional:\n")
|
|
b.WriteString("- When the user tells you to remember something, or shares a preference, fact, decision, or how-to worth keeping, save it with mem_add (or mem_edit to update an existing page) — do it on your own, without being asked.\n")
|
|
b.WriteString("- Before doing any task or answering, first call mem_search to see whether your memory already holds the answer or how to do it, then mem_read the best page. Do this even when the request has new specifics like a name, a place or a value — your saved method still applies, only the parameter changes.\n")
|
|
case MemOnDemand:
|
|
b.WriteString("\nMemory is ON-DEMAND: you have the mem_* tools but do NOT read or write memory on your own. Call mem_search/mem_read only when the user explicitly asks you to recall or look something up, and mem_add/mem_edit only when the user explicitly asks you to remember something. Otherwise leave memory untouched and answer directly.\n")
|
|
}
|
|
if caps.Agent {
|
|
// Règle courte mais indispensable : sans elle le modèle fabrique ses scripts
|
|
// avec echo/python -c, ce que cmd.exe massacre (guillemets imbriqués).
|
|
b.WriteString("\nTo create or rewrite a file, always use write (or edit to patch one) — never echo, cat, python -c or Set-Content through the shell; quoting breaks. Write the script with write, then run it with bash.\n")
|
|
b.WriteString("For anything about the system or files, use bash instead of guessing. Act immediately — call the right tool, then answer. Never end your turn after only thinking. Be concise.\n")
|
|
if caps.Mem == MemAlways {
|
|
b.WriteString("Before answering any question about yourself or this machine, always call mem_search first — even trivial-seeming ones. Testing with a tool never replaces this: memory may hold context the tool won't reveal. Search memory, then verify, then answer.\n")
|
|
}
|
|
}
|
|
if caps.Internet {
|
|
b.WriteString("\nWeb access (Crawl4AI): web_search (DuckDuckGo), web_open (fetch a URL → metadata + outline), web_read (read a line range of an opened URL), web_grep (regex in an opened URL). Workflow: web_open first, then web_read/web_grep.\n")
|
|
year := time.Now().Format("2006")
|
|
b.WriteString("Your training data is stale. For ANY question about recent/latest/current things (releases, versions, news, prices, scores, fixtures, 'since when') call web_search BEFORE writing any date or version. Your answer must match the dates/facts you actually read.\n")
|
|
b.WriteString("SEARCH QUERY YEAR RULE: today is in " + year + ". If your query includes a year, use ONLY " + year + " — NEVER write a past year like " + prevYear(year) + " that you remember from training; it silently biases results toward stale pages. Default: put no year at all and let the freshest result win. Don't hedge ('probably', 'I think') about a fact a tool can verify — search instead.\n")
|
|
}
|
|
if caps.Agent {
|
|
if line := mcpPromptLine(); line != "" {
|
|
b.WriteString(line)
|
|
}
|
|
}
|
|
b.WriteString("\nDate: " + time.Now().Format("2006-01-02"))
|
|
return b.String()
|
|
}
|
|
|
|
// prevYear returns the year before the given "2006"-formatted year string, used
|
|
// to name explicitly the stale year the model must NOT put in search queries.
|
|
func prevYear(year string) string {
|
|
n, err := strconv.Atoi(year)
|
|
if err != nil {
|
|
return year
|
|
}
|
|
return strconv.Itoa(n - 1)
|
|
}
|
|
|
|
// machineSystemPrompt returns a short briefing about the host the model is
|
|
// running on, so that when machine access is enabled it knows *which* machine
|
|
// run_shell acts upon (and doesn't claim it has no access to "your PC").
|
|
// Returns "" when machine access is off.
|
|
func machineSystemPrompt(caps Caps) string {
|
|
if !caps.Agent {
|
|
return ""
|
|
}
|
|
host, _ := os.Hostname()
|
|
if host == "" {
|
|
host = "unknown"
|
|
}
|
|
who := ""
|
|
if u, err := user.Current(); err == nil {
|
|
who = u.Username
|
|
}
|
|
cwd := agentWorkspace()
|
|
|
|
var b strings.Builder
|
|
b.WriteString(fmt.Sprintf("Machine: host=%s, %s/%s", host, runtime.GOOS, runtime.GOARCH))
|
|
if who != "" {
|
|
b.WriteString(", user=" + who)
|
|
}
|
|
if cwd != "" {
|
|
b.WriteString(", cwd=" + cwd)
|
|
}
|
|
b.WriteString(".")
|
|
if cwd != "" {
|
|
b.WriteString(" Relative paths in write/edit/bash resolve inside this working folder — put scratch files there. Write outside it ONLY with an absolute path the user explicitly asked for; never scatter files into the folder jean was launched from.")
|
|
}
|
|
return b.String()
|
|
}
|
|
|
|
// runShell executes a command via the platform shell (bash -c on Unix, cmd /C
|
|
// on Windows — see newShellCmd in sys_platform_*.go) with a clamped timeout,
|
|
// returning a single string formatted "exit: N\n\nstdout:\n...\n\nstderr:\n..."
|
|
// truncated to keep tool output bounded.
|
|
func runShell(command string, timeoutSec int) string {
|
|
if timeoutSec <= 0 {
|
|
timeoutSec = toolDefaultTimeout
|
|
}
|
|
if timeoutSec > toolMaxTimeout {
|
|
timeoutSec = toolMaxTimeout
|
|
}
|
|
ctx, cancel := context.WithTimeout(context.Background(), time.Duration(timeoutSec)*time.Second)
|
|
defer cancel()
|
|
cmd := newShellCmd(ctx, command)
|
|
// Le shell démarre dans le workspace, pas dans le dossier d'où jean a été
|
|
// lancé : un `> notes.txt` du modèle ne doit pas atterrir sur le Bureau.
|
|
cmd.Dir = agentWorkspace()
|
|
var stdout, stderr strings.Builder
|
|
cmd.Stdout = &stdout
|
|
cmd.Stderr = &stderr
|
|
err := cmd.Run()
|
|
if ctx.Err() == context.DeadlineExceeded {
|
|
return fmt.Sprintf("[timeout après %ds]", timeoutSec)
|
|
}
|
|
exit := 0
|
|
if err != nil {
|
|
if ee, ok := err.(*exec.ExitError); ok {
|
|
exit = ee.ExitCode()
|
|
} else {
|
|
return fmt.Sprintf("[erreur: %v]", err)
|
|
}
|
|
}
|
|
out := tailRunes(stdout.String(), toolMaxOutput)
|
|
errOut := tailRunes(stderr.String(), toolMaxOutput)
|
|
parts := []string{fmt.Sprintf("exit: %d", exit)}
|
|
if out != "" {
|
|
parts = append(parts, "stdout:\n"+out)
|
|
}
|
|
if errOut != "" {
|
|
parts = append(parts, "stderr:\n"+errOut)
|
|
}
|
|
return strings.Join(parts, "\n\n")
|
|
}
|
|
|
|
// shellName is the shell runShell actually spawns on this platform. The model is
|
|
// told this explicitly: advertising the tool as "bash" on Windows made it emit
|
|
// bash quoting into cmd.exe, which mangles it (unterminated string literals, and
|
|
// stray "Commande ECHO activée." landing inside generated files).
|
|
func shellName() string {
|
|
if runtime.GOOS == "windows" {
|
|
return "cmd.exe"
|
|
}
|
|
return "bash"
|
|
}
|
|
|
|
// fileWrite writes content to path verbatim, creating parent directories and
|
|
// replacing any existing file. This is the escape hatch from shell quoting: a
|
|
// model with only a shell has to build files with echo/python -c, which is
|
|
// unreliable everywhere and outright broken on cmd.exe.
|
|
func fileWrite(path, content string) string {
|
|
if strings.TrimSpace(path) == "" {
|
|
return "[erreur] chemin vide"
|
|
}
|
|
path = resolveAgentPath(path)
|
|
if dir := filepath.Dir(path); dir != "" && dir != "." {
|
|
if err := os.MkdirAll(dir, 0o755); err != nil {
|
|
return "[erreur] " + err.Error()
|
|
}
|
|
}
|
|
// Préserve les permissions d'origine quand le fichier existe déjà (un script
|
|
// 0755 réécrit doit rester exécutable).
|
|
mode := os.FileMode(0o644)
|
|
existed := false
|
|
if fi, err := os.Stat(path); err == nil {
|
|
mode = fi.Mode()
|
|
existed = true
|
|
}
|
|
if err := os.WriteFile(path, []byte(content), mode); err != nil {
|
|
return "[erreur] " + err.Error()
|
|
}
|
|
verb := "créé"
|
|
if existed {
|
|
verb = "réécrit"
|
|
}
|
|
return fmt.Sprintf("[ok] %s %s (%d octets)", path, verb, len(content))
|
|
}
|
|
|
|
// fileEdit applies a single exact-text replacement to a file on disk: oldText
|
|
// must appear EXACTLY once (otherwise it errors), so the model can patch a file
|
|
// without rewriting it whole. Returns a short status string for the tool result.
|
|
func fileEdit(path, oldText, newText string) string {
|
|
if strings.TrimSpace(path) == "" {
|
|
return "[erreur] chemin vide"
|
|
}
|
|
if oldText == "" {
|
|
return "[erreur] old vide"
|
|
}
|
|
path = resolveAgentPath(path)
|
|
b, err := os.ReadFile(path)
|
|
if err != nil {
|
|
return "[erreur] " + err.Error()
|
|
}
|
|
content := string(b)
|
|
n := strings.Count(content, oldText)
|
|
if n == 0 {
|
|
// Modification déjà en place : on le dit clairement plutôt que de renvoyer
|
|
// une erreur, sinon le modèle croit avoir échoué et recommence.
|
|
if newText != "" && strings.Contains(content, newText) {
|
|
return "[ok] déjà à jour — le fichier contient déjà cette modification"
|
|
}
|
|
return "[erreur] old introuvable dans le fichier"
|
|
}
|
|
if n > 1 {
|
|
return fmt.Sprintf("[erreur] old apparaît %d fois — ajoute du contexte pour le rendre unique", n)
|
|
}
|
|
updated := strings.Replace(content, oldText, newText, 1)
|
|
// Préserve les permissions d'origine (un script 0755 doit rester exécutable).
|
|
mode := os.FileMode(0o644)
|
|
if fi, err := os.Stat(path); err == nil {
|
|
mode = fi.Mode()
|
|
}
|
|
if err := os.WriteFile(path, []byte(updated), mode); err != nil {
|
|
return "[erreur] " + err.Error()
|
|
}
|
|
return fmt.Sprintf("[ok] %s modifié (1 remplacement)", path)
|
|
}
|
|
|
|
// tailRunes returns the last n runes of s (used to cap tool output).
|
|
func tailRunes(s string, n int) string {
|
|
r := []rune(s)
|
|
if len(r) <= n {
|
|
return s
|
|
}
|
|
return string(r[len(r)-n:])
|
|
}
|