@trac3er/oh-my-god 2.2.3 → 2.2.4

package/build/lib/hooks/_token_counter.py

@@ -1,221 +0,0 @@
-#!/usr/bin/env python3
-"""Tiered token estimation helpers for OMG hooks."""
-from __future__ import annotations
-
-import json
-import os
-import importlib
-import urllib.request
-from collections.abc import Iterable
-
-API_URL = "https://api.anthropic.com/v1/messages/count_tokens"
-API_MODEL = "claude-3-5-haiku-20241022"
-
-_FEATURE_UI_DISPLAY = "ui_display"
-_FEATURE_BUDGET_ENFORCEMENT = "budget_enforcement"
-_FEATURE_PREFLIGHT = "preflight"
-
-
-def _safe_int(value: float) -> int:
-    if value <= 0:
-        return 0
-    return int(value)
-
-
-def _extract_features(text: str) -> tuple[int, int, int]:
-    encoded = text.encode("utf-8")
-    byte_count = len(encoded)
-    word_count = len(text.split())
-    line_count = text.count("\n") + (1 if text else 0)
-    return byte_count, word_count, line_count
-
-
-def _default_coefficients() -> tuple[float, float, float, float]:
-    return (1.0, 0.19, 0.75, 1.1)
-
-
-def _gaussian_solve(matrix: list[list[float]], vector: list[float]) -> list[float] | None:
-    n = len(vector)
-    if n == 0:
-        return None
-    try:
-        for i in range(n):
-            pivot = i
-            for r in range(i + 1, n):
-                if abs(matrix[r][i]) > abs(matrix[pivot][i]):
-                    pivot = r
-            if abs(matrix[pivot][i]) < 1e-12:
-                return None
-
-            if pivot != i:
-                matrix[i], matrix[pivot] = matrix[pivot], matrix[i]
-                vector[i], vector[pivot] = vector[pivot], vector[i]
-
-            pivot_val = matrix[i][i]
-            for c in range(i, n):
-                matrix[i][c] /= pivot_val
-            vector[i] /= pivot_val
-
-            for r in range(n):
-                if r == i:
-                    continue
-                factor = matrix[r][i]
-                if factor == 0:
-                    continue
-                for c in range(i, n):
-                    matrix[r][c] -= factor * matrix[i][c]
-                vector[r] -= factor * vector[i]
-        return vector
-    except Exception:
-        return None
-
-
-def _fit_linear_coefficients(samples: Iterable[tuple[str, int]]) -> tuple[float, float, float, float]:
-    rows: list[list[float]] = []
-    targets: list[float] = []
-    for text, tokens in samples:
-        bcount, wcount, lcount = _extract_features(text)
-        rows.append([1.0, float(bcount), float(wcount), float(lcount)])
-        targets.append(float(tokens))
-
-    dim = 4
-    if len(rows) < dim:
-        return _default_coefficients()
-
-    xtx = [[0.0 for _ in range(dim)] for _ in range(dim)]
-    xty = [0.0 for _ in range(dim)]
-    for row, target in zip(rows, targets):
-        for i in range(dim):
-            xty[i] += row[i] * target
-            for j in range(dim):
-                xtx[i][j] += row[i] * row[j]
-
-    solved = _gaussian_solve(xtx, xty)
-    if solved is None:
-        return _default_coefficients()
-
-    return (float(solved[0]), float(solved[1]), float(solved[2]), float(solved[3]))
-
-
-_CALIBRATION_SAMPLES: tuple[tuple[str, int], ...] = (
-    ("ls", 5),
-    ("git status", 7),
-    ("echo 'hello world'", 10),
-    ("python3 -m pytest tests/hooks/test_feature_flags_v2.py -q", 18),
-    ("def hello():\n    return 'world'\n", 24),
-    (
-        """from pathlib import Path\nfor path in Path('hooks').glob('*.py'):\n    print(path.name)\n""",
-        44,
-    ),
-    (
-        """def estimate_tokens(text: str, tier: int = 1) -> int:\n    if tier == 1:\n        return max(1, int(len(text) / 3.5))\n    return 0\n""",
-        80,
-    ),
-    (
-        "\n".join(["line with common code and comments" for _ in range(80)]),
-        720,
-    ),
-    (
-        "\n".join(["longer source line with punctuation () {} [] == != <= >=" for _ in range(250)]),
-        1800,
-    ),
-)
-
-_COEFFICIENTS = _fit_linear_coefficients(_CALIBRATION_SAMPLES)
-
-
-def _estimate_tier1(text: str) -> int:
-    if not text:
-        return 0
-    return max(1, int(len(text) / 3.5))
-
-
-def _estimate_tier2(text: str) -> int:
-    if not text:
-        return 0
-    bias, w_bytes, w_words, w_lines = _COEFFICIENTS
-    bcount, wcount, lcount = _extract_features(text)
-    prediction = bias + (w_bytes * bcount) + (w_words * wcount) + (w_lines * lcount)
-    return max(1, _safe_int(prediction))
-
-
-def _get_anthropic_api_key() -> str | None:
-    try:
-        store_mod = importlib.import_module("credential_store")
-        key = store_mod.get_active_key("anthropic")
-        if key:
-            return key
-    except (ImportError, RuntimeError, ValueError, OSError, AttributeError):
-        pass
-    return os.environ.get("ANTHROPIC_API_KEY")
-
-
-def _estimate_tier3(text: str) -> int:
-    if not text:
-        return 0
-
-    api_key = _get_anthropic_api_key()
-    if not api_key:
-        return _estimate_tier2(text)
-
-    payload = {
-        "model": API_MODEL,
-        "messages": [{"role": "user", "content": text}],
-    }
-    body = json.dumps(payload).encode("utf-8")
-    request = urllib.request.Request(
-        API_URL,
-        data=body,
-        method="POST",
-        headers={
-            "content-type": "application/json",
-            "x-api-key": api_key,
-            "anthropic-version": "2023-06-01",
-        },
-    )
-
-    try:
-        with urllib.request.urlopen(request, timeout=8) as response:
-            raw = response.read().decode("utf-8")
-        parsed = json.loads(raw)
-        token_value = parsed.get("input_tokens")
-        if isinstance(token_value, int) and token_value >= 0:
-            return token_value
-    except Exception:
-        return _estimate_tier2(text)
-
-    return _estimate_tier2(text)
-
-
-def auto_select_tier(operation: str, text: str = "") -> int:
-    normalized = (operation or "").strip().lower()
-    if normalized == _FEATURE_UI_DISPLAY:
-        return 1
-    if normalized == _FEATURE_BUDGET_ENFORCEMENT:
-        return 2
-    if normalized == _FEATURE_PREFLIGHT:
-        if len(text) >= 8000:
-            return 3
-        if _estimate_tier2(text) >= 1000:
-            return 3
-        return 2
-    return 1
-
-
-def estimate_tokens(text: str, tier: int = 1) -> int:
-    """Estimate token count with 3 reliability/cost tiers.
-
-    Tier 1: fast heuristic (`len(text)/3.5`).
-    Tier 2: calibrated linear model using bytes, words, lines.
-    Tier 3: Anthropic count_tokens API with graceful fallback to tier 2.
-    """
-    try:
-        if tier == 1:
-            return _estimate_tier1(text)
-        if tier == 2:
-            return _estimate_tier2(text)
-        if tier == 3:
-            return _estimate_tier3(text)
-        return _estimate_tier1(text)
-    except Exception:
-        return _estimate_tier1(text)

package/build/lib/hooks/branch_manager.py

@@ -1,253 +0,0 @@
-#!/usr/bin/env python3
-"""SessionStart Hook — Smart Branch Manager.
-
-Auto-creates a feature branch when on main/master/develop.
-Extracts task description from OMG state files for branch naming.
-
-Feature-gated: OMG_GIT_WORKFLOW_ENABLED (uses get_feature_flag('GIT_WORKFLOW'))
-"""
-import json
-import os
-import re
-import subprocess
-import sys
-from datetime import datetime
-
-HOOKS_DIR = os.path.dirname(__file__)
-if HOOKS_DIR not in sys.path:
-    sys.path.insert(0, HOOKS_DIR)
-
-from _common import setup_crash_handler, json_input, get_feature_flag
-
-setup_crash_handler("branch-manager", fail_closed=False)
-
-# Default branches that trigger feature branch creation
-DEFAULT_BRANCHES = frozenset({"main", "master", "develop"})
-
-# Max length for the descriptive part of branch name
-MAX_BRANCH_NAME_LEN = 50
-
-
-def _get_project_dir() -> str:
-    """Get project directory from env or cwd."""
-    return os.environ.get("CLAUDE_PROJECT_DIR", os.getcwd())
-
-
-def _has_git(project_dir: str) -> bool:
-    """Check if project_dir is inside a git repo."""
-    try:
-        result = subprocess.run(
-            ["git", "-C", project_dir, "rev-parse", "--git-dir"],
-            capture_output=True, text=True, timeout=5,
-        )
-        return result.returncode == 0
-    except Exception:
-        return False
-
-
-def _current_branch(project_dir: str) -> str | None:
-    """Get current branch name. Returns None on failure."""
-    try:
-        result = subprocess.run(
-            ["git", "-C", project_dir, "rev-parse", "--abbrev-ref", "HEAD"],
-            capture_output=True, text=True, timeout=5,
-        )
-        if result.returncode == 0:
-            return result.stdout.strip()
-    except Exception:
-        pass
-    return None
-
-
-def _sanitize_branch_name(description: str) -> str:
-    """Sanitize a description into a valid git branch name segment.
-
-    Rules:
-    - Lowercase
-    - Replace spaces/underscores with hyphens
-    - Strip special characters (keep alphanumeric and hyphens)
-    - Collapse consecutive hyphens
-    - Strip leading/trailing hyphens
-    - Max MAX_BRANCH_NAME_LEN chars
-    """
-    name = description.lower().strip()
-    # Replace spaces and underscores with hyphens
-    name = re.sub(r"[\s_]+", "-", name)
-    # Remove everything except alphanumeric and hyphens
-    name = re.sub(r"[^a-z0-9-]", "", name)
-    # Collapse consecutive hyphens
-    name = re.sub(r"-{2,}", "-", name)
-    # Strip leading/trailing hyphens
-    name = name.strip("-")
-    # Truncate to max length, but don't cut mid-word if possible
-    if len(name) > MAX_BRANCH_NAME_LEN:
-        truncated = name[:MAX_BRANCH_NAME_LEN]
-        # Try to cut at last hyphen to avoid mid-word truncation
-        last_hyphen = truncated.rfind("-")
-        if last_hyphen > 20:
-            truncated = truncated[:last_hyphen]
-        name = truncated.rstrip("-")
-    return name
-
-
-def _extract_task_description(project_dir: str) -> str | None:
-    """Extract task description from OMG state files.
-
-    Priority order:
-    (a) .omg/state/_plan.md title (first # heading)
-    (b) .omg/state/_checklist.md first item
-    (c) .omg/state/working-memory.md last entry
-    (d) fallback: None (caller uses session-{timestamp})
-    """
-    state_dir = os.path.join(project_dir, ".omg", "state")
-
-    # (a) Plan title
-    plan_path = os.path.join(state_dir, "_plan.md")
-    if os.path.isfile(plan_path):
-        try:
-            with open(plan_path, "r", encoding="utf-8", errors="ignore") as f:
-                for line in f:
-                    line = line.strip()
-                    if line.startswith("# "):
-                        title = line[2:].strip()
-                        if title:
-                            return title
-        except Exception:
-            pass
-
-    # (b) Checklist first item
-    checklist_path = os.path.join(state_dir, "_checklist.md")
-    if os.path.isfile(checklist_path):
-        try:
-            with open(checklist_path, "r", encoding="utf-8", errors="ignore") as f:
-                for line in f:
-                    line = line.strip()
-                    # Match markdown checkbox items: - [ ] or - [x]
-                    m = re.match(r"^-\s*\[.\]\s*(.+)$", line)
-                    if m:
-                        item = m.group(1).strip()
-                        if item:
-                            return item
-        except Exception:
-            pass
-
-    # (c) Working memory last entry
-    wm_path = os.path.join(state_dir, "working-memory.md")
-    if os.path.isfile(wm_path):
-        try:
-            with open(wm_path, "r", encoding="utf-8", errors="ignore") as f:
-                content = f.read()
-            # Split by ## headings, take last entry's content
-            sections = re.split(r"\n## ", content)
-            if len(sections) > 1:
-                # Last section: first line is heading, rest is content
-                last_lines = sections[-1].split("\n")
-                # Find first non-empty content line after heading
-                for line in last_lines[1:]:
-                    line = line.strip()
-                    if line:
-                        return line
-                # Fallback to heading if no content
-                heading = last_lines[0].strip()
-                if heading:
-                    return heading
-            elif sections:
-                # Single section — try first non-empty non-heading line
-                for line in sections[0].split("\n"):
-                    line = line.strip()
-                    if line and not line.startswith("#"):
-                        return line
-        except Exception:
-            pass
-
-    # (d) No state files found
-    return None
-
-
-def _is_merge_writer_locked(project_dir: str) -> bool:
-    try:
-        _root = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
-        if _root not in sys.path:
-            sys.path.insert(0, _root)
-        from runtime.merge_writer import MergeWriter  # pyright: ignore[reportMissingImports]
-        return MergeWriter(project_dir).is_locked()
-    except Exception:
-        return False
-
-
-def _create_branch(project_dir: str, branch_name: str) -> bool:
-    """Create and checkout a new branch. Returns True on success."""
-    if _is_merge_writer_locked(project_dir):
-        print(
-            f"[OMG branch-manager] Branch creation blocked: merge-writer lock is held",
-            file=sys.stderr,
-        )
-        return False
-    try:
-        result = subprocess.run(
-            ["git", "-C", project_dir, "checkout", "-b", branch_name],
-            capture_output=True, text=True, timeout=10,
-        )
-        return result.returncode == 0
-    except Exception:
-        return False
-
-
-def main() -> None:
-    """Main hook entry point."""
-    data = json_input()
-
-    # Feature gate: exit silently if disabled
-    if not get_feature_flag("GIT_WORKFLOW", default=False):
-        sys.exit(0)
-
-    project_dir = _get_project_dir()
-
-    # No-op if not a git repo
-    if not _has_git(project_dir):
-        sys.exit(0)
-
-    # Get current branch
-    branch = _current_branch(project_dir)
-    if branch is None:
-        sys.exit(0)
-
-    # No-op if already on a non-default branch (feature branch, etc.)
-    if branch not in DEFAULT_BRANCHES:
-        sys.exit(0)
-
-    # Extract task description and build branch name
-    description = _extract_task_description(project_dir)
-    if description:
-        sanitized = _sanitize_branch_name(description)
-    else:
-        sanitized = ""
-
-    if not sanitized:
-        # Fallback: session-{timestamp}
-        sanitized = f"session-{datetime.now().strftime('%Y%m%d-%H%M%S')}"
-
-    target_branch = f"feature/{sanitized}"
-
-    # Dry-run mode: output what would happen without executing
-    dry_run = os.environ.get("OMG_GIT_WORKFLOW_DRY_RUN", "").lower() in ("1", "true", "yes")
-    if dry_run:
-        print(
-            f"[OMG branch-manager] DRY-RUN: Would create branch '{target_branch}' from '{branch}'",
-            file=sys.stderr,
-        )
-        sys.exit(0)
-
-    # Create the feature branch
-    success = _create_branch(project_dir, target_branch)
-    if success:
-        print(
-            f"[OMG branch-manager] Created branch '{target_branch}' from '{branch}'",
-            file=sys.stderr,
-        )
-
-    sys.exit(0)
-
-
-if __name__ == "__main__":
-    main()