intelligence-suite 0.2.4__tar.gz → 0.2.6__tar.gz

{intelligence_suite-0.2.4 → intelligence_suite-0.2.6}/MentorIntelligence/content/ingest_practices.py

@@ -95,18 +95,66 @@ def _parse_text_practice(file: Path, root: Path) -> list[dict]:
     if len(text) < 30:
         return []
 
-    heading    = stem.replace("_", " ").replace("-", " ").title()
-    chunk_text = (
-        f"Practice: {heading} (from {file.name})\n"
-        f"Category: general\n"
-        f"---\n{text[:3000]}"
-    )
-    return [make_chunk(
-        domain="mentor", type_="practice",
-        locator=f"practice.{_sanitize(stem)}",
-        text=chunk_text, source=rel, language="markdown",
-        metadata={"category": "general", "priority": "normal", "audience": ["all"]},
-    )]
+    # Extract H1 as document-level title for context
+    doc_title = stem.replace("_", " ").replace("-", " ").title()
+    for line in text.splitlines():
+        if line.startswith("# ") and not line.startswith("## "):
+            doc_title = line[2:].strip()
+            break
+
+    # Split by H2 headings → one chunk per section
+    sections = _split_by_h2(text, doc_title)
+
+    chunks = []
+    for section_title, section_body in sections:
+        chunk_text = (
+            f"Practice: {doc_title} — {section_title} (from {file.name})\n"
+            f"Category: general\n"
+            f"---\n{section_body.strip()[:4000]}"
+        )
+        chunks.append(make_chunk(
+            domain="mentor", type_="practice",
+            locator=f"practice.{_sanitize(stem)}.{_sanitize(section_title)}",
+            text=chunk_text, source=rel, language="markdown",
+            metadata={
+                "category": "general", "priority": "normal", "audience": ["all"],
+                "parent_title": doc_title, "section": section_title,
+            },
+        ))
+    return chunks
+
+
+def _split_by_h2(text: str, doc_title: str) -> list[tuple[str, str]]:
+    """Split markdown by H2 headings. Returns [(section_title, body), ...]."""
+    lines = text.splitlines()
+    sections: list[tuple[str, str]] = []
+    current_title = doc_title          # content before first H2 → intro section
+    current_body: list[str] = []
+
+    for line in lines:
+        if line.startswith("# ") and not line.startswith("## "):
+            # H1 = document title, skip (already captured above)
+            continue
+        if line.startswith("## "):
+            # Flush previous section
+            body = "\n".join(current_body).strip()
+            if len(body) >= 20:
+                sections.append((current_title, body))
+            current_title = line[3:].strip()
+            current_body = []
+        else:
+            current_body.append(line)
+
+    # Last section
+    body = "\n".join(current_body).strip()
+    if len(body) >= 20:
+        sections.append((current_title, body))
+
+    # Fallback: whole file as one chunk if no H2 found
+    if not sections:
+        sections = [(doc_title, text[:4000])]
+
+    return sections
 
 
 def _parse_yaml_practice(file: Path, root: Path) -> list[dict]:

intelligence_suite-0.2.6/MentorIntelligence/content/path_templates.json

@@ -0,0 +1,174 @@
+{
+  "developer": {
+    "description": "Percorso per developer che entrano su un progetto esistente",
+    "steps": [
+      {
+        "id": "dev_01",
+        "title": "Architettura generale",
+        "description": "Capire la struttura del progetto e i componenti principali",
+        "sources": ["doc", "code"],
+        "suggested_queries": [
+          "Qual è l'architettura del progetto?",
+          "Quali sono i componenti principali?"
+        ],
+        "checkpoint": "Sai descrivere i 3 componenti principali e le loro responsabilità"
+      },
+      {
+        "id": "dev_02",
+        "title": "Setup ambiente di sviluppo",
+        "description": "Configurare l'ambiente locale e installare le dipendenze",
+        "sources": ["doc"],
+        "suggested_queries": [
+          "Come si configura l'ambiente di sviluppo?",
+          "Quali dipendenze servono?"
+        ],
+        "checkpoint": "Hai il progetto che gira in locale"
+      },
+      {
+        "id": "dev_03",
+        "title": "Autenticazione e sicurezza",
+        "description": "Capire il sistema di autenticazione e le politiche di sicurezza",
+        "sources": ["code", "doc"],
+        "suggested_queries": [
+          "Come funziona l'autenticazione?",
+          "Dove viene validato il token?"
+        ],
+        "checkpoint": "Sai dove viene gestita l'auth e come testarla"
+      },
+      {
+        "id": "dev_04",
+        "title": "Prassi di sviluppo",
+        "description": "Conoscere le convenzioni del team: git, review, naming",
+        "sources": ["mentor"],
+        "suggested_queries": [
+          "Naming convention branch",
+          "Come si fa una code review?"
+        ],
+        "checkpoint": "Conosci le regole del team su git, review, deploy"
+      },
+      {
+        "id": "dev_05",
+        "title": "Deploy e operazioni",
+        "description": "Capire la procedura di deploy e rollback",
+        "sources": ["doc"],
+        "suggested_queries": [
+          "Come si fa il deploy?",
+          "Cosa fare se qualcosa va storto?"
+        ],
+        "checkpoint": "Sai leggere la procedura di deploy e rollback"
+      }
+    ]
+  },
+  "non_developer": {
+    "description": "Percorso per ruoli non tecnici (PM, HR, sales, operations)",
+    "steps": [
+      {
+        "id": "nd_01",
+        "title": "Cos'è il progetto e a cosa serve",
+        "description": "Capire il prodotto e il suo valore di business",
+        "sources": ["doc"],
+        "suggested_queries": [
+          "Cosa fa IntelligenceSuite?",
+          "A chi è rivolto?"
+        ],
+        "checkpoint": "Sai spiegare il prodotto in 3 frasi a un non-tecnico"
+      },
+      {
+        "id": "nd_02",
+        "title": "Processi e prassi aziendali",
+        "description": "Conoscere i processi e le prassi del team",
+        "sources": ["mentor", "doc"],
+        "suggested_queries": [
+          "Quali sono le prassi del team?",
+          "Come si prendono le decisioni?"
+        ],
+        "checkpoint": "Conosci i processi principali del team"
+      },
+      {
+        "id": "nd_03",
+        "title": "Come interagire con il team tecnico",
+        "description": "Sapere come comunicare e collaborare con i developer",
+        "sources": ["mentor"],
+        "suggested_queries": [
+          "Come si apre una richiesta?",
+          "Chi contattare per cosa?"
+        ],
+        "checkpoint": "Sai a chi rivolgerti per i casi principali"
+      }
+    ]
+  },
+  "mixed": {
+    "description": "Percorso ibrido: tecnico con responsabilità trasversali",
+    "steps": [
+      {
+        "id": "dev_01",
+        "title": "Architettura generale",
+        "description": "Capire la struttura del progetto",
+        "sources": ["doc", "code"],
+        "suggested_queries": ["Qual è l'architettura del progetto?"],
+        "checkpoint": "Sai descrivere i componenti principali"
+      },
+      {
+        "id": "nd_02",
+        "title": "Processi e prassi aziendali",
+        "description": "Conoscere i processi del team",
+        "sources": ["mentor", "doc"],
+        "suggested_queries": ["Quali sono le prassi del team?"],
+        "checkpoint": "Conosci i processi principali"
+      },
+      {
+        "id": "dev_03",
+        "title": "Autenticazione e sicurezza",
+        "description": "Capire il sistema di autenticazione",
+        "sources": ["code", "doc"],
+        "suggested_queries": ["Come funziona l'autenticazione?"],
+        "checkpoint": "Sai dove viene gestita l'auth"
+      },
+      {
+        "id": "dev_04",
+        "title": "Prassi di sviluppo",
+        "description": "Convenzioni del team tecnico",
+        "sources": ["mentor"],
+        "suggested_queries": ["Naming convention branch", "Code review process"],
+        "checkpoint": "Conosci le regole del team"
+      },
+      {
+        "id": "nd_03",
+        "title": "Come interagire con il team tecnico",
+        "description": "Comunicazione e collaborazione cross-team",
+        "sources": ["mentor"],
+        "suggested_queries": ["Come si apre una richiesta?"],
+        "checkpoint": "Sai a chi rivolgerti"
+      },
+      {
+        "id": "dev_05",
+        "title": "Deploy e operazioni",
+        "description": "Capire deploy e rollback",
+        "sources": ["doc"],
+        "suggested_queries": ["Come si fa il deploy?"],
+        "checkpoint": "Sai leggere la procedura di deploy"
+      }
+    ]
+  },
+  "unknown": {
+    "description": "Percorso generico per utenti con profilo non ancora rilevato",
+    "steps": [
+      {
+        "id": "nd_01",
+        "title": "Cos'è il progetto e a cosa serve",
+        "description": "Panoramica generale del prodotto",
+        "sources": ["doc"],
+        "suggested_queries": ["Cosa fa questo progetto?"],
+        "checkpoint": "Hai capito lo scopo del progetto"
+      },
+      {
+        "id": "nd_02",
+        "title": "Processi e prassi aziendali",
+        "description": "Processi del team",
+        "sources": ["mentor", "doc"],
+        "suggested_queries": ["Quali sono le prassi del team?"],
+        "checkpoint": "Conosci i processi principali"
+      }
+    ]
+  }
+}

{intelligence_suite-0.2.4 → intelligence_suite-0.2.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: intelligence-suite
-Version: 0.2.4
+Version: 0.2.6
 Summary: Modular knowledge retrieval suite: code, docs, APIs — all on-premise
 Author-email: ViciusLio <viciuslios@gmail.com>
 License-Expression: MIT
@@ -491,8 +491,8 @@ codebase and company documents via cross-domain retrieval.
 ### CLI quickstart
 
 ```bash
-# (Optional) ingest best practices
-mkdir practices && echo "# Git conventions\n..." > practices/git.md
+# Ingest your team's best practices (Markdown / TXT / YAML)
+# The repo ships with a ready-to-use practices/ folder for IntelligenceSuite itself:
 mi-ingest ./practices
 
 # Start the mentor server (default: http://localhost:8082)
@@ -509,6 +509,21 @@ curl -X POST http://localhost:8082/api/v1/mentor/ask \
   -d '{"session_id": "...", "question": "How does authentication work in this codebase?"}'
 ```
 
+### Bundled `practices/` folder
+
+The repository ships with four ready-to-use Markdown guides covering IntelligenceSuite
+itself — ideal for teams adopting the suite:
+
+| File | Content |
+|---|---|
+| `01_onboarding_nuovo_developer.md` | Day-by-day setup, first indexing, team conventions |
+| `02_come_usare_code_intelligence.md` | CI pipeline, CLI commands, API, troubleshooting |
+| `03_come_usare_doc_intelligence.md` | DI supported formats, pipeline, confidence notes |
+| `04_come_usare_mentor_intelligence.md` | MI onboarding flow, profile detection, REST API |
+
+Each file is split by `##` headings at ingest time — **one chunk per section** — for
+precise retrieval. Running `mi-ingest ./practices` produces ~30 chunks.
+
 ---
 
 ## What a chunk looks like

{intelligence_suite-0.2.4 → intelligence_suite-0.2.6}/README.md

@@ -432,8 +432,8 @@ codebase and company documents via cross-domain retrieval.
 ### CLI quickstart
 
 ```bash
-# (Optional) ingest best practices
-mkdir practices && echo "# Git conventions\n..." > practices/git.md
+# Ingest your team's best practices (Markdown / TXT / YAML)
+# The repo ships with a ready-to-use practices/ folder for IntelligenceSuite itself:
 mi-ingest ./practices
 
 # Start the mentor server (default: http://localhost:8082)
@@ -450,6 +450,21 @@ curl -X POST http://localhost:8082/api/v1/mentor/ask \
   -d '{"session_id": "...", "question": "How does authentication work in this codebase?"}'
 ```
 
+### Bundled `practices/` folder
+
+The repository ships with four ready-to-use Markdown guides covering IntelligenceSuite
+itself — ideal for teams adopting the suite:
+
+| File | Content |
+|---|---|
+| `01_onboarding_nuovo_developer.md` | Day-by-day setup, first indexing, team conventions |
+| `02_come_usare_code_intelligence.md` | CI pipeline, CLI commands, API, troubleshooting |
+| `03_come_usare_doc_intelligence.md` | DI supported formats, pipeline, confidence notes |
+| `04_come_usare_mentor_intelligence.md` | MI onboarding flow, profile detection, REST API |
+
+Each file is split by `##` headings at ingest time — **one chunk per section** — for
+precise retrieval. Running `mi-ingest ./practices` produces ~30 chunks.
+
 ---
 
 ## What a chunk looks like

{intelligence_suite-0.2.4 → intelligence_suite-0.2.6}/intelligence_suite.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: intelligence-suite
-Version: 0.2.4
+Version: 0.2.6
 Summary: Modular knowledge retrieval suite: code, docs, APIs — all on-premise
 Author-email: ViciusLio <viciuslios@gmail.com>
 License-Expression: MIT
@@ -491,8 +491,8 @@ codebase and company documents via cross-domain retrieval.
 ### CLI quickstart
 
 ```bash
-# (Optional) ingest best practices
-mkdir practices && echo "# Git conventions\n..." > practices/git.md
+# Ingest your team's best practices (Markdown / TXT / YAML)
+# The repo ships with a ready-to-use practices/ folder for IntelligenceSuite itself:
 mi-ingest ./practices
 
 # Start the mentor server (default: http://localhost:8082)
@@ -509,6 +509,21 @@ curl -X POST http://localhost:8082/api/v1/mentor/ask \
   -d '{"session_id": "...", "question": "How does authentication work in this codebase?"}'
 ```
 
+### Bundled `practices/` folder
+
+The repository ships with four ready-to-use Markdown guides covering IntelligenceSuite
+itself — ideal for teams adopting the suite:
+
+| File | Content |
+|---|---|
+| `01_onboarding_nuovo_developer.md` | Day-by-day setup, first indexing, team conventions |
+| `02_come_usare_code_intelligence.md` | CI pipeline, CLI commands, API, troubleshooting |
+| `03_come_usare_doc_intelligence.md` | DI supported formats, pipeline, confidence notes |
+| `04_come_usare_mentor_intelligence.md` | MI onboarding flow, profile detection, REST API |
+
+Each file is split by `##` headings at ingest time — **one chunk per section** — for
+precise retrieval. Running `mi-ingest ./practices` produces ~30 chunks.
+
 ---
 
 ## What a chunk looks like

{intelligence_suite-0.2.4 → intelligence_suite-0.2.6}/intelligence_suite.egg-info/SOURCES.txt

@@ -30,6 +30,7 @@ MentorIntelligence/profile_detector.py
 MentorIntelligence/session_manager.py
 MentorIntelligence/content/__init__.py
 MentorIntelligence/content/ingest_practices.py
+MentorIntelligence/content/path_templates.json
 intelligence_core/__init__.py
 intelligence_core/chunk.py
 intelligence_core/config.py

{intelligence_suite-0.2.4 → intelligence_suite-0.2.6}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "intelligence-suite"
-version = "0.2.4"
+version = "0.2.6"
 description = "Modular knowledge retrieval suite: code, docs, APIs — all on-premise"
 readme = "README.md"
 license = "MIT"
@@ -72,6 +72,9 @@ Author     = "https://www.linkedin.com/in/vincenzo-lio-dataengineer/"
 where = ["."]
 include = ["intelligence_core*", "CodeIntelligence*", "DocIntelligence*", "MentorIntelligence*", "intelligence_ui*"]
 
+[tool.setuptools.package-data]
+"MentorIntelligence.content" = ["*.json"]
+
 [tool.ruff]
 line-length = 100
 target-version = "py310"