⚡️ Speed up function speedup_critic by 23% in PR #620 (pre-commit-update)

The optimization introduces a **cached GitHub Actions detection mechanism** that eliminates repeated expensive environment lookups in the hot path.

**Key optimization**: 
- Added `_in_github_actions_mode()` function with `@lru_cache(maxsize=1)` that caches the result of `env_utils.get_pr_number()` 
- Replaced the inline `bool(env_utils.get_pr_number())` call inside `speedup_critic` with a call to the cached function

**Why this is faster**:
The original code called `env_utils.get_pr_number()` on every invocation of `speedup_critic`. While `get_pr_number()` itself is cached, it still involves function call overhead and the boolean conversion. The line profiler shows this operation took **5.69ms** (28.8% of total time) in the original vs only **1.09ms** (7.7%) in the optimized version.

**Performance characteristics**:
- **Best for high-frequency scenarios**: The optimization shines when `speedup_critic` is called many times (as shown in the large-scale tests with 500+ candidates), where the 22% speedup compounds significantly
- **Minimal impact on single calls**: For individual calls, the improvement is small but consistent
- **GitHub Actions environments benefit most**: Since the cached result prevents repeated environment variable lookups that are expensive in CI environments

The optimization maintains identical behavior while reducing redundant work through strategic caching of the GitHub Actions detection logic.

Author: codeflash-ai[bot]

codeflash/code_utils/env_utils.py

@@ -8,6 +8,7 @@
 from typing import Any, Optional
 
 from codeflash.cli_cmds.console import console, logger
+from codeflash.code_utils import env_utils
 from codeflash.code_utils.code_utils import exit_with_message
 from codeflash.code_utils.formatter import format_code
 from codeflash.code_utils.shell_utils import read_api_key_from_shell_config
@@ -129,3 +130,8 @@ def is_pr_draft() -> bool:
     """Check if the PR is draft. in the github action context."""
     event = get_cached_gh_event_data()
     return bool(event.get("pull_request", {}).get("draft", False))
+
+
+@lru_cache(maxsize=1)
+def _in_github_actions_mode() -> bool:
+    return bool(env_utils.get_pr_number())

codeflash/result/critic.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+from functools import lru_cache
 from typing import TYPE_CHECKING
 
 from codeflash.cli_cmds.console import logger
@@ -9,7 +10,7 @@
     MIN_IMPROVEMENT_THRESHOLD,
     MIN_TESTCASE_PASSED_THRESHOLD,
 )
-from codeflash.models.models import TestType
+from codeflash.models.models import OptimizedCandidateResult, TestType
 
 if TYPE_CHECKING:
     from codeflash.models.models import CoverageData, OptimizedCandidateResult, OriginalCodeBaseline
@@ -41,8 +42,7 @@ def speedup_critic(
     """
     noise_floor = 3 * MIN_IMPROVEMENT_THRESHOLD if original_code_runtime < 10000 else MIN_IMPROVEMENT_THRESHOLD
     if not disable_gh_action_noise:
-        in_github_actions_mode = bool(env_utils.get_pr_number())
-        if in_github_actions_mode:
+        if _in_github_actions_mode():
             noise_floor = noise_floor * 2  # Increase the noise floor in GitHub Actions mode
 
     perf_gain = performance_gain(
@@ -76,3 +76,8 @@ def coverage_critic(original_code_coverage: CoverageData | None, test_framework:
     if original_code_coverage:
         return original_code_coverage.coverage >= COVERAGE_THRESHOLD
     return False
+
+
+@lru_cache(maxsize=1)
+def _in_github_actions_mode() -> bool:
+    return bool(env_utils.get_pr_number())