codedna 0.9.1__tar.gz → 0.9.2__tar.gz

{codedna-0.9.1 → codedna-0.9.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: codedna
-Version: 0.9.1
+Version: 0.9.2
 Summary: AI Code Transparency Tool — detect AI-written code, analyze commit quality, track understanding debt, sprint health, bus factor, and more
 Project-URL: Homepage, https://codedna.dev
 Project-URL: Repository, https://github.com/natureco-official/codedna
@@ -44,10 +44,13 @@ Requires-Dist: uvicorn[standard]>=0.49.0
 Provides-Extra: dashboard
 Requires-Dist: next-cmd>=0.5.0; extra == 'dashboard'
 Provides-Extra: dev
-Requires-Dist: black>=24.0.0; extra == 'dev'
-Requires-Dist: pytest-cov>=4.0.0; extra == 'dev'
-Requires-Dist: pytest>=8.0.0; extra == 'dev'
-Requires-Dist: ruff>=0.1.0; extra == 'dev'
+Requires-Dist: black>=26.5.1; extra == 'dev'
+Requires-Dist: pytest-cov>=7.1.0; extra == 'dev'
+Requires-Dist: pytest>=9.1.1; extra == 'dev'
+Requires-Dist: ruff>=0.15.19; extra == 'dev'
+Provides-Extra: nl
+Requires-Dist: huggingface-hub; extra == 'nl'
+Requires-Dist: pyarrow<18,>=14; extra == 'nl'
 Description-Content-Type: text/markdown
 
 <div align="center">

{codedna-0.9.1 → codedna-0.9.2}/codedna/analyzer.py

@@ -51,11 +51,9 @@ try:
     import tree_sitter_typescript as tstypescript
     _TS_LANG = tstypescript.language_typescript()
     _TSX_LANG = tstypescript.language_tsx()
-    _TS_AVAILABLE = True
 except Exception:
     _TS_LANG = tsjavascript.language()
     _TSX_LANG = tsjavascript.language()
-    _TS_AVAILABLE = False
 
 # Supported language map
 LANGUAGE_MAP: dict[str, tuple] = {

{codedna-0.9.1 → codedna-0.9.2}/codedna/api.py

@@ -69,7 +69,7 @@ app = FastAPI(
 # CORS — fully open for dashboard or external tools
 app.add_middleware(
     CORSMiddleware,
-    allow_origins=["*"],
+    allow_origins=["http://localhost:3000"],
     allow_methods=["*"],
     allow_headers=["*"],
 )

{codedna-0.9.1 → codedna-0.9.2}/codedna/cli.py

@@ -2,6 +2,7 @@
 
 from __future__ import annotations
 
+import os
 import sys
 from datetime import datetime
 from pathlib import Path
@@ -499,7 +500,6 @@ def serve(
     repo: Optional[Path] = typer.Option(None, "--repo", "-r", help="Git repo directory"),
 ) -> None:
     """Start the FastAPI REST server."""
-    import os
     import uvicorn
 
     root = repo or find_git_root() or Path.cwd()
@@ -813,7 +813,7 @@ def pr_comment(
     """Post CodeDNA analysis comment to GitHub PR (updates existing comment, no spam)."""
     import os
     from codedna.integrations.github_bot import (
-        format_pr_comment, post_or_update_comment, github_actions_pr_infosi,
+        format_pr_comment, post_or_update_comment, github_actions_pr_bilgisi,
     )
     from codedna.tech_debt import calculate_repo_debt
 
@@ -830,7 +830,7 @@ def pr_comment(
     target_pr = pr
 
     if not target_repo or not target_pr:
-        info = github_actions_pr_infosi()
+        info = github_actions_pr_bilgisi()
         if info:
             auto_repo, auto_pr = info
             target_repo = target_repo or auto_repo
@@ -1130,13 +1130,13 @@ def interview_list(
     table.add_column("Notes", min_width=20)
 
     for o in sessions:
-        score_str = f"{o['skor']:.1f}/5" if o["skor"] is not None else "[dim]—[/dim]"
+        score_str = f"{o['score']:.1f}/5" if o["score"] is not None else "[dim]—[/dim]"
         table.add_row(
             str(o["id"]),
-            o["candidate"] or o.get("aday") or "?",
-            o["start_time"] or o.get("start_date") or "?",
+            o["candidate"] or "?",
+            o["start_time"] or "?",
             score_str,
-            (o["notes"] or o.get("notlar") or "")[:30],
+            (o["notes"] or "")[:30],
         )
 
     console.print()
@@ -1188,7 +1188,7 @@ def bus_factor(
     from codedna.bus_factor import calculate_bus_factor, get_at_risk_files, _LARGE_REPO_THRESHOLD
     from codedna.plan import is_feature_available
 
-        # Plan check
+    # Plan check
     if not is_feature_available("bus_factor"):
         console.print(
             Panel(
@@ -2041,6 +2041,21 @@ def doctor(
     console.print()
 
 
+# ---------------------------------------------------------------------------
+# codedna ask (natural language → command, powered by Needle)
+# ---------------------------------------------------------------------------
+
+
+@app.command()
+def ask(
+    query: list[str] = typer.Argument(..., help="What you want to do, in plain language."),
+    dry_run: bool = typer.Option(False, "--dry-run", help="Show the command without running it."),
+) -> None:
+    """Run a CodeDNA command from a natural-language request (powered by Needle)."""
+    from codedna.nl import route_command
+    route_command(" ".join(query), dry_run=dry_run)
+
+
 # ---------------------------------------------------------------------------
 # codedna update (self-upgrade from PyPI)
 # ---------------------------------------------------------------------------

{codedna-0.9.1 → codedna-0.9.2}/codedna/db.py

@@ -147,6 +147,7 @@ def save_file_score(
     complexity_score: float,
     comment_ratio: float,
     understanding_score: Optional[float] = None,
+    ai_tool_guess: Optional[str] = None,
     db_path: Optional[Path] = None,
 ) -> None:
     """Save a file analysis score."""
@@ -154,10 +155,10 @@ def save_file_score(
         conn.execute(
             """
             INSERT INTO file_scores
-                (commit_hash, file_path, ai_probability, complexity_score, comment_ratio, understanding_score)
-            VALUES (?, ?, ?, ?, ?, ?)
+                (commit_hash, file_path, ai_probability, complexity_score, comment_ratio, understanding_score, ai_tool_guess)
+            VALUES (?, ?, ?, ?, ?, ?, ?)
             """,
-            (commit_hash, file_path, ai_probability, complexity_score, comment_ratio, understanding_score),
+            (commit_hash, file_path, ai_probability, complexity_score, comment_ratio, understanding_score, ai_tool_guess),
         )
 
 
@@ -227,8 +228,13 @@ def get_all_file_understanding_scores(
             FROM file_scores fs
             JOIN commits c ON fs.commit_hash = c.commit_hash
             WHERE fs.understanding_score IS NOT NULL
-            GROUP BY fs.file_path
-            HAVING c.timestamp = MAX(c.timestamp)
+              AND c.timestamp = (
+                  SELECT MAX(c2.timestamp)
+                  FROM file_scores fs2
+                  JOIN commits c2 ON fs2.commit_hash = c2.commit_hash
+                  WHERE fs2.file_path = fs.file_path
+                    AND fs2.understanding_score IS NOT NULL
+              )
             """,
         ).fetchall()
     return {r["file_path"]: float(r["understanding_score"]) for r in rows}

codedna-0.9.2/codedna/nl.py

@@ -0,0 +1,234 @@
+"""codedna/nl.py — Natural language router for CodeDNA, powered by Needle.
+
+Lets a user type a request in plain language and have it translated
+into the matching `codedna` CLI invocation. The model itself is
+multilingual at the input level, so the command works regardless of
+which language the user types in — only the code, comments, and
+output messages are kept in English for consistency with the rest
+of the project.
+
+Usage (standalone):
+    python -m codedna.nl "show me the AI score for my last commit"
+
+Usage (from cli.py):
+    from codedna.nl import route_command
+    route_command(user_query)
+"""
+
+from __future__ import annotations
+
+import json
+import subprocess
+import sys
+from pathlib import Path
+
+from needle import SimpleAttentionNetwork, load_checkpoint, generate, get_tokenizer
+
+# ---------------------------------------------------------------------------
+# Tool definitions — mirrors codedna_tools.json. Keep these two in sync.
+# ---------------------------------------------------------------------------
+
+TOOLS_PATH = Path(__file__).parent / "nl_tools.json"
+
+
+def _load_tools() -> str:
+    if TOOLS_PATH.exists():
+        return TOOLS_PATH.read_text(encoding="utf-8")
+    raise FileNotFoundError(
+        f"{TOOLS_PATH} not found. Copy codedna_tools.json into codedna/ as nl_tools.json."
+    )
+
+
+# ---------------------------------------------------------------------------
+# Needle model (loaded once, lazily)
+# ---------------------------------------------------------------------------
+
+# Checkpoint paths. The fine-tuned checkpoint (if present) takes priority
+# over the base Hugging Face download. Both are kept under ~/.codedna/needle/
+# so they are shared across projects and never committed by accident.
+_CHECKPOINT_DIR = Path.home() / ".codedna" / "needle"
+_FINETUNED_PATH = _CHECKPOINT_DIR / "needle-finetuned.pkl"
+_CHECKPOINT_PATH = _CHECKPOINT_DIR / "needle.pkl"
+_HF_REPO_ID = "Cactus-Compute/needle"
+_HF_FILENAME = "needle.pkl"
+
+_MODEL = None
+_PARAMS = None
+_TOKENIZER = None
+
+
+def _resolve_checkpoint() -> Path:
+    """Return the best available checkpoint path.
+
+    Priority:
+      1. Fine-tuned checkpoint at ~/.codedna/needle/needle-finetuned.pkl
+      2. Already-downloaded base model at ~/.codedna/needle/needle.pkl
+      3. Auto-downloaded from Hugging Face, cached at (2)
+    """
+    if _FINETUNED_PATH.exists():
+        return _FINETUNED_PATH
+    if _CHECKPOINT_PATH.exists():
+        return _CHECKPOINT_PATH
+
+    try:
+        from huggingface_hub import hf_hub_download
+    except ImportError as exc:
+        raise ImportError(
+            "huggingface_hub is required to auto-download the Needle checkpoint. "
+            "Install it with: pip install codedna[nl]"
+        ) from exc
+
+    print("First run: downloading the Needle model (~53 MB)...", file=sys.stderr)
+    _CHECKPOINT_DIR.mkdir(parents=True, exist_ok=True)
+    downloaded_path = hf_hub_download(repo_id=_HF_REPO_ID, filename=_HF_FILENAME)
+
+    # hf_hub_download caches under ~/.cache/huggingface; copy it into our own
+    # checkpoint dir so the path stays stable and predictable for the CLI.
+    import shutil
+    shutil.copyfile(downloaded_path, _CHECKPOINT_PATH)
+    print(f"Needle model cached at {_CHECKPOINT_PATH}", file=sys.stderr)
+    return _CHECKPOINT_PATH
+
+
+def _ensure_model_loaded() -> None:
+    global _MODEL, _PARAMS, _TOKENIZER
+    if _MODEL is not None:
+        return
+    checkpoint_path = _resolve_checkpoint()
+    params, config = load_checkpoint(str(checkpoint_path))
+    _MODEL = SimpleAttentionNetwork(config)
+    _PARAMS = params
+    _TOKENIZER = get_tokenizer()
+
+
+# ---------------------------------------------------------------------------
+# Mapping from tool call -> actual `codedna` CLI argv
+# ---------------------------------------------------------------------------
+
+def _build_argv(name: str, args: dict) -> list[str]:
+    """Translate a Needle tool call into a `codedna ...` argv list."""
+
+    def opt(flag: str, key: str, cast=str) -> list[str]:
+        if key in args and args[key] not in (None, ""):
+            return [flag, str(cast(args[key]))]
+        return []
+
+    def flag(name_flag: str, key: str) -> list[str]:
+        return [name_flag] if args.get(key) else []
+
+    mapping = {
+        "scan": ["scan", *opt("--max", "max_files", int), *opt("--min-risk", "min_risk", float)],
+        "status": ["status"],
+        "commit_history": ["history", *opt("--limit", "limit", int)],
+        "bus_factor": [
+            "bus-factor",
+            *flag("--critical", "critical"),
+            *opt("--max", "max_files", int),
+        ],
+        "debt": ["debt", *opt("--rate", "rate", float), *opt("--file", "file")],
+        "protect_add": [
+            "protect", "add", str(args.get("file_path", "")),
+            *opt("--threshold", "threshold", float),
+            *opt("--label", "label"),
+        ],
+        "protect_remove": ["protect", "remove", str(args.get("file_path", ""))],
+        "protect_list": ["protect", "list"],
+        "protect_check": ["protect", "check"],
+        "sprint_create": [
+            "sprint", "create",
+            "--name", str(args.get("name", "")),
+            "--start", str(args.get("start", "")),
+            "--end", str(args.get("end", "")),
+        ],
+        "sprint_health": ["sprint", "health"],
+        "sprint_history": ["sprint", "history", *opt("--limit", "limit", int)],
+        "export": ["export", *opt("--format", "fmt"), *opt("--output", "output")],
+        "webhook_show": ["webhook", "--show"],
+        "webhook_test": ["webhook", "--test"],
+        "webhook_reset": ["webhook", "--reset"],
+        "watch": [
+            "watch",
+            *opt("--interval", "interval", int),
+            *flag("--once", "once"),
+            *flag("--notify", "notify"),
+        ],
+        "report": ["report"],
+        "doctor": ["doctor"],
+    }
+
+    if name not in mapping:
+        raise ValueError(f"Unknown tool name returned by Needle: {name!r}")
+    return mapping[name]
+
+
+# Commands that change state. At the current model accuracy, a misrouted
+# request here (e.g. "protect" resolving to protect_remove instead of
+# protect_add) can do the opposite of what the user asked. These always
+# require an explicit confirmation, regardless of --dry-run.
+_MUTATING_TOOLS = {
+    "protect_add",
+    "protect_remove",
+    "sprint_create",
+    "webhook_reset",
+}
+
+
+# ---------------------------------------------------------------------------
+# Public entrypoint
+# ---------------------------------------------------------------------------
+
+def route_command(user_query: str, dry_run: bool = False) -> None:
+    """Parse `user_query` with Needle and run the matching codedna command(s)."""
+    _ensure_model_loaded()
+
+    raw = generate(
+        _MODEL, _PARAMS, _TOKENIZER,
+        query=user_query,
+        tools=_load_tools(),
+        stream=False,
+    )
+
+    try:
+        calls = json.loads(raw)
+    except json.JSONDecodeError:
+        print(f"Unexpected output from Needle: {raw}", file=sys.stderr)
+        return
+
+    if not calls:
+        print("Could not match that request to a known command. Try rephrasing.")
+        return
+
+    for call in calls:
+        name = call.get("name")
+        args = call.get("arguments", {})
+        try:
+            argv = _build_argv(name, args)
+        except ValueError as exc:
+            print(str(exc), file=sys.stderr)
+            continue
+
+        full_cmd = ["codedna", *argv]
+        print(f"-> {' '.join(full_cmd)}")
+
+        if dry_run:
+            continue
+
+        if name in _MUTATING_TOOLS:
+            # Needle is a small (26M parameter) model and can misroute
+            # requests, especially the difference between "protect" and
+            # "unprotect" intents. Never run a state-changing command
+            # silently — always get an explicit yes from the user first.
+            answer = input("Run this command? [y/N] ").strip().lower()
+            if answer not in ("y", "yes"):
+                print("Skipped.")
+                continue
+
+        subprocess.run(full_cmd)
+
+
+if __name__ == "__main__":
+    query = " ".join(sys.argv[1:])
+    if not query:
+        print("Usage: python -m codedna.nl '<your request in natural language>'")
+        sys.exit(1)
+    route_command(query)

codedna-0.9.2/codedna/nl_tools.json

@@ -0,0 +1,127 @@
+[
+  {
+    "name": "scan",
+    "description": "Scan the current repo and show an AI-generated code risk report.",
+    "parameters": {
+      "max_files": {"type": "integer", "description": "Maximum number of files to scan. Defaults to 200.", "required": false},
+      "min_risk": {"type": "number", "description": "Minimum AI probability filter, 0.0 to 1.0.", "required": false}
+    }
+  },
+  {
+    "name": "status",
+    "description": "Show the AI score and commit message quality for the last commit.",
+    "parameters": {}
+  },
+  {
+    "name": "commit_history",
+    "description": "Show a table of past commits with their AI scores and understanding scores. This is about commit history, not sprint history.",
+    "parameters": {
+      "limit": {"type": "integer", "description": "Number of past commits to show. Defaults to 20.", "required": false}
+    }
+  },
+  {
+    "name": "bus_factor",
+    "description": "Run repo-wide bus factor analysis and show ownership risks.",
+    "parameters": {
+      "critical": {"type": "boolean", "description": "Show only critical files where bus factor equals 1.", "required": false},
+      "max_files": {"type": "integer", "description": "Maximum number of files to process. Defaults to 500.", "required": false}
+    }
+  },
+  {
+    "name": "debt",
+    "description": "Calculate technical debt cost in dollars for the whole repository. Only set the file parameter if the user names a specific file path explicitly.",
+    "parameters": {
+      "rate": {"type": "number", "description": "Hourly developer rate in dollars. Defaults to 75.", "required": false},
+      "file": {"type": "string", "description": "An actual file path, e.g. 'src/auth.py'. Do not fill this with a description or topic — only a real path the user typed.", "required": false}
+    }
+  },
+  {
+    "name": "protect_add",
+    "description": "Add protection to a file or directory so AI is not allowed to edit it. Use this when the user wants to START protecting, guard, or lock a path. This is the opposite of protect_remove.",
+    "parameters": {
+      "file_path": {"type": "string", "description": "Path to protect, e.g. 'auth/' or 'payment/login.py'.", "required": true},
+      "threshold": {"type": "number", "description": "Minimum required understanding score, 0 to 5. Defaults to 3.5.", "required": false},
+      "label": {"type": "string", "description": "Human-readable label for this protected module.", "required": false}
+    }
+  },
+  {
+    "name": "protect_remove",
+    "description": "Remove protection from a previously protected file, so AI is allowed to edit it again. Use this when the user wants to STOP, unprotect, lift protection, or remove the protected status from a path. This is the opposite of protect_add.",
+    "parameters": {
+      "file_path": {"type": "string", "description": "Path to unprotect.", "required": true}
+    }
+  },
+  {
+    "name": "protect_list",
+    "description": "Show all protected modules and their current status.",
+    "parameters": {}
+  },
+  {
+    "name": "protect_check",
+    "description": "Show only protected modules that currently violate their understanding threshold.",
+    "parameters": {}
+  },
+  {
+    "name": "sprint_create",
+    "description": "Create a new sprint and calculate its health score.",
+    "parameters": {
+      "name": {"type": "string", "description": "Sprint name, e.g. 'Sprint 14'.", "required": true},
+      "start": {"type": "string", "description": "Start date in YYYY-MM-DD format.", "required": true},
+      "end": {"type": "string", "description": "End date in YYYY-MM-DD format.", "required": true}
+    }
+  },
+  {
+    "name": "sprint_health",
+    "description": "Show the latest sprint's health score, velocity, and quality.",
+    "parameters": {}
+  },
+  {
+    "name": "sprint_history",
+    "description": "Show past sprints as a table.",
+    "parameters": {
+      "limit": {"type": "integer", "description": "Number of past sprints to show. Defaults to 10.", "required": false}
+    }
+  },
+  {
+    "name": "export",
+    "description": "Export all CodeDNA data as JSON or CSV.",
+    "parameters": {
+      "fmt": {"type": "string", "description": "Export format: 'json' or 'csv'. Defaults to json.", "required": false},
+      "output": {"type": "string", "description": "Output file path. If omitted, prints to stdout.", "required": false}
+    }
+  },
+  {
+    "name": "webhook_show",
+    "description": "Show the current Slack or Discord webhook configuration.",
+    "parameters": {}
+  },
+  {
+    "name": "webhook_test",
+    "description": "Send a test webhook notification to Slack or Discord.",
+    "parameters": {}
+  },
+  {
+    "name": "webhook_reset",
+    "description": "Clear the current webhook configuration.",
+    "parameters": {}
+  },
+  {
+    "name": "watch",
+    "description": "Watch the git repo for new commits and analyze them automatically.",
+    "parameters": {
+      "interval": {"type": "integer", "description": "Poll interval in seconds. Defaults to 30.", "required": false},
+      "once": {"type": "boolean", "description": "Run once and exit, useful for cron jobs.", "required": false},
+      "notify": {"type": "boolean", "description": "Send a webhook notification when a new commit is found.", "required": false}
+    }
+  },
+  {
+    "name": "report",
+    "description": "Generate an HTML report for the repo.",
+    "parameters": {}
+  },
+  {
+    "name": "doctor",
+    "description": "Run a system health check on the CodeDNA installation itself (dependencies, config, database). This is about the CodeDNA tool's own health, not sprint health or project health.",
+    "parameters": {}
+  }
+]

{codedna-0.9.1 → codedna-0.9.2}/codedna/scorer.py

@@ -74,7 +74,8 @@ def calculate_single_commit_ratio(repo: Repo, file_path: str) -> float:
         largest = max(commit_contributions)
         return min(largest / total, 1.0)
 
-    except Exception:
+    except Exception as exc:
+        console.log(f"[dim]single_commit_ratio error for {file_path}: {exc}[/dim]")
         return 0.0
 
 

{codedna-0.9.1 → codedna-0.9.2}/codedna/tech_debt.py

@@ -128,7 +128,7 @@ def calculate_file_debt(
         try:
             from codedna.analyzer import analyze_file
             result = analyze_file(p)
-            if result.desteklenmiyor or result.hata:
+            if result.unsupported or result.error:
                 return None
             ai = result.ai_probability
             complexity = result.complexity_score

{codedna-0.9.1 → codedna-0.9.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "codedna"
-version = "0.9.1"
+version = "0.9.2"
 description = "AI Code Transparency Tool — detect AI-written code, analyze commit quality, track understanding debt, sprint health, bus factor, and more"
 readme = "README.md"
 requires-python = ">=3.10"
@@ -50,11 +50,16 @@ dependencies = [
 dashboard = [
     "next-cmd>=0.5.0",  # Next.js process management
 ]
+nl = [
+    # "needle @ git+https://github.com/cactus-compute/needle.git",  # TODO: publish needle to PyPI first
+    "huggingface_hub",
+    "pyarrow>=14,<18",
+]
 dev = [
-    "pytest>=8.0.0",
-    "pytest-cov>=4.0.0",
-    "black>=24.0.0",
-    "ruff>=0.1.0",
+    "pytest>=9.1.1",
+    "pytest-cov>=7.1.0",
+    "black>=26.5.1",
+    "ruff>=0.15.19",
 ]
 
 [project.urls]
@@ -66,6 +71,9 @@ Changelog = "https://github.com/natureco-official/codedna/blob/main/CHANGELOG.md
 [project.scripts]
 codedna = "codedna.cli:main"
 
+[tool.hatch.metadata]
+allow-direct-references = true
+
 [build-system]
 requires = ["hatchling"]
 build-backend = "hatchling.build"