@allansantos-dev/smart-tool 0.9.37 → 0.9.38

package/CHANGELOG.md

@@ -1,5 +1,15 @@
 # Changelog
 
+## 0.9.38 - beta
+
+- One definition of a search term. `smart_search mode=dossier` now takes its terms from `dossier.focus_terms`, the
+  function `smart_scout` already used for its windows (`scout.focus_terms` is the same function): words of 3+ characters,
+  question words dropped, camelCase split. A note is no longer ranked by "why" or "the", an identifier like
+  `SessionStart` finds the note that writes "session start", and a query made only of question words is refused.
+  Words of 2 characters no longer count in dossier queries. Ranking cases in `tests/test_dossier.py`.
+- One copy of the hook-entry removal in `client_hooks.py`: the session-start audit entry and the redirect entries share
+  `_ours` and `_without_ours`, which take the predicate that says which hooks are ours.
+
 ## 0.9.37 - beta
 
 - The site no longer sends a human reader to the repository. The "Release notes" links (hero and footer) go to a

package/README.md

@@ -16,7 +16,7 @@ registered as the `playwright` server where this machine lets a browser be drive
 An optional hook routes the agent's native Grep/Glob/Read/Bash/WebSearch/WebFetch calls to these tools when they
 are cheaper, or just tells the agent what Smart Tool would do.
 
-**Status:** 0.9.37 beta. Windows only. What changed in each version: [CHANGELOG.md](CHANGELOG.md).
+**Status:** 0.9.38 beta. Windows only. What changed in each version: [CHANGELOG.md](CHANGELOG.md).
 
 ## How it works
 

package/client_hooks.py

@@ -58,18 +58,18 @@ def _entries(client):
     return entries
 
 
-def _ours(group):
-    return isinstance(group, dict) and isinstance(group.get("hooks"), list) and any(map(_our_hook, group["hooks"]))
+def _ours(group, is_ours=_our_hook):
+    return isinstance(group, dict) and isinstance(group.get("hooks"), list) and any(map(is_ours, group["hooks"]))
 
 
-def _without_ours(groups):
+def _without_ours(groups, is_ours=_our_hook):
     """Drops only our hook entries; a group shared with other hooks keeps them and disappears only when emptied."""
     kept = []
     for group in groups:
-        if not _ours(group):
+        if not _ours(group, is_ours):
             kept.append(group)
             continue
-        others = [hook for hook in group["hooks"] if not _our_hook(hook)]
+        others = [hook for hook in group["hooks"] if not is_ours(hook)]
         if others:
             kept.append({**group, "hooks": others})
     return kept
@@ -276,10 +276,6 @@ def _audit_hook(hook):
     return isinstance(hook, dict) and AUDIT_MARKER in str(hook.get("command", ""))
 
 
-def _audit_ours(group):
-    return isinstance(group, dict) and isinstance(group.get("hooks"), list) and any(map(_audit_hook, group["hooks"]))
-
-
 def _audit_entry():
     return {"matcher": AUDIT_MATCHER,
             "hooks": [{"type": "command", "command": f"{_shell_safe(sys.executable)} {_shell_safe(AUDIT_SCRIPT)}",
@@ -299,15 +295,7 @@ def _audit_plan():
     groups = data["hooks"].setdefault("SessionStart", [])
     if not isinstance(groups, list):
         raise ValueError(f"{path}: hooks.SessionStart is not a list; nothing was changed.")
-    kept = []
-    for group in groups:
-        if not _audit_ours(group):
-            kept.append(group)
-            continue
-        others = [hook for hook in group["hooks"] if not _audit_hook(hook)]
-        if others:
-            kept.append({**group, "hooks": others})
-    groups[:] = kept
+    groups[:] = _without_ours(groups, _audit_hook)
     return path, data, groups
 
 
@@ -321,7 +309,7 @@ def audit_state():
     except FileNotFoundError:
         return "missing"
     groups = ((data.get("hooks") or {}).get("SessionStart") or []) if isinstance(data, dict) else []
-    found = [group for group in groups if _audit_ours(group)] if isinstance(groups, list) else []
+    found = [group for group in groups if _ours(group, _audit_hook)] if isinstance(groups, list) else []
     if not found:
         return "missing"
     return "current" if _audit_entry() in found else "outdated"

package/dossier.py

@@ -16,13 +16,29 @@ TITLE_CHARS = 80
 MAX_FILE_BYTES = 512 * 1024
 BACKLOG_FILES = ("docs/BACKLOG.md", "BACKLOG.md")
 INSTRUCTION_FILES = (("CLAUDE.md", "claude_md"), (".claude/CLAUDE.md", "claude_md"), ("AGENTS.md", "agents_md"))
-_TERM = re.compile(r"[\w]{2,}", re.UNICODE)
+_WORD = re.compile(r"[\w]{3,}", re.UNICODE)
+_STOPWORDS = {
+    "the", "and", "for", "how", "what", "where", "which", "does", "with", "from", "that", "this", "are", "when", "who",
+    "why", "into", "used", "use", "uses", "its", "has", "have", "code", "project", "file", "files",
+    "como", "onde", "qual", "quais", "que", "quem", "por", "para", "com", "uma", "dos", "das", "nos", "nas", "pelo",
+    "pela", "ele", "ela", "isso", "esse", "essa", "este", "esta", "fica", "ficam", "usa", "usado", "usada", "sobre",
+    "entre", "depois", "antes", "quando", "cada", "seu", "sua", "seus", "suas", "não", "sem", "mais", "código",
+    "projeto", "arquivo", "arquivos", "decide", "funciona",
+}
 _UNIT_START = re.compile(r"#{1,6}\s|[-*]\s")
 
 
-def _terms(query):
-    """Up to 24 distinct casefolded word terms of the query."""
-    return list(dict.fromkeys(term.casefold() for term in _TERM.findall(query or "")))[:24]
+def focus_terms(*texts):
+    """Casefolded content words of the texts (3+ characters, common question words dropped), camelCase split, in
+    order: what a window is scored by."""
+    found = []
+    for text in texts:
+        spaced = re.sub(r"([a-z])([A-Z])", r"\1 \2", text or "")
+        for word in _WORD.findall(spaced):
+            folded = word.casefold()
+            if folded not in _STOPWORDS and folded not in found:
+                found.append(folded)
+    return found[:24]
 
 
 def _sources(root):
@@ -78,7 +94,7 @@ def query(root, text, timeout_ms=DEFAULT_TIMEOUT_MS, max_hits=DEFAULT_MAX_HITS):
     was found until then."""
     started = time.monotonic()
     deadline = started + timeout_ms / 1000
-    terms = _terms(text)
+    terms = focus_terms(text)
     reasons, scanned, unreadable, scored = [], 0, [], []
     for rel, kind in _sources(root):
         if time.monotonic() > deadline:
@@ -116,7 +132,7 @@ def answer(arguments):
         raise ValueError("smart_search mode=dossier requires 'project_root' pointing to an existing folder")
     text = next((arguments[key] for key in ("query_identifiers", "query_comments", "doc_query")
                  if isinstance(arguments.get(key), str) and arguments[key].strip()), "")
-    if not _terms(text):
+    if not focus_terms(text):
         raise ValueError("smart_search mode=dossier requires 'query_identifiers' (or 'doc_query') with at least one word.")
     timeout_ms = _bounded(arguments, "timeout_ms", DEFAULT_TIMEOUT_MS, 1, MAX_TIMEOUT_MS)
     max_hits = _bounded(arguments, "max_hits", DEFAULT_MAX_HITS, 1, MAX_HITS)

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@allansantos-dev/smart-tool",
-  "version": "0.9.37",
+  "version": "0.9.38",
   "description": "Local MCP server that gives coding agents (Claude Code, Codex) cheaper, sharper tools than their built-in search. Windows.",
   "license": "Apache-2.0",
   "author": "Allan Santos",

package/scout.py

@@ -12,11 +12,11 @@ Market pattern copied (docs/research/2026-10-24-scout-modelo-leve-no-smart-tool.
 excerpts, never an agent prompt, so no agent context is reopened per call.
 """
 import os
-import re
 
 import document_text
 import index_scope
 import indexer
+from dossier import focus_terms  # noqa: F401  (one definition of a search term, shared with smart_search mode=dossier)
 
 MAX_QUESTION_CHARS = 2000
 MAX_TERMS_CHARS = 400
@@ -35,15 +35,6 @@ PLAN_MAX_FOLDERS = 6
 OUTLINE_MAX_DIRS = 300
 OUTLINE_MAX_ROOT_FILES = 200
 _FIELDS = {"project_root", "question", "terms", "max_files"}
-_WORD = re.compile(r"[\w]{3,}", re.UNICODE)
-_STOPWORDS = {
-    "the", "and", "for", "how", "what", "where", "which", "does", "with", "from", "that", "this", "are", "when", "who",
-    "why", "into", "used", "use", "uses", "its", "has", "have", "code", "project", "file", "files",
-    "como", "onde", "qual", "quais", "que", "quem", "por", "para", "com", "uma", "dos", "das", "nos", "nas", "pelo",
-    "pela", "ele", "ela", "isso", "esse", "essa", "este", "esta", "fica", "ficam", "usa", "usado", "usada", "sobre",
-    "entre", "depois", "antes", "quando", "cada", "seu", "sua", "seus", "suas", "não", "sem", "mais", "código",
-    "projeto", "arquivo", "arquivos", "decide", "funciona",
-}
 
 PLAN_SYSTEM = (
     "You plan a keyword scan of a code project that has no search index yet. Given a question and the project's "
@@ -100,19 +91,6 @@ def arguments(raw):
     return {"project_root": root, "question": question.strip(), "terms": (terms or "").strip(), "max_files": max_files}
 
 
-def focus_terms(*texts):
-    """Casefolded content words of the texts (3+ characters, common question words dropped), camelCase split, in
-    order: what a window is scored by."""
-    found = []
-    for text in texts:
-        spaced = re.sub(r"([a-z])([A-Z])", r"\1 \2", text or "")
-        for word in _WORD.findall(spaced):
-            folded = word.casefold()
-            if folded not in _STOPWORDS and folded not in found:
-                found.append(folded)
-    return found[:24]
-
-
 def disk_lines(root, path, start, end):
     """`[(line_number, text)]` of lines `start..end` of the file as it is on disk now, or None when the file cannot be
     read by the index rules (secret, dotfile, too large, binary, gone) or holds no line text (docx)."""

package/version.py

@@ -1,7 +1,7 @@
 """Single source of the Smart Tool version and of the User-Agent sent to public APIs."""
 import re
 
-VERSION = "0.9.37"
+VERSION = "0.9.38"
 _CONTACT_RE = re.compile(r"^(?:[^\s@()<>;]+@[^\s@()<>;]+\.[^\s@()<>;]+|https?://[^\s()<>;]+)$")