sourcecode 1.31.30__tar.gz → 1.31.31__tar.gz

{sourcecode-1.31.30 → sourcecode-1.31.31}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sourcecode
-Version: 1.31.30
+Version: 1.31.31
 Summary: Deterministic codebase context for AI coding agents
 License:                                  Apache License
                                    Version 2.0, January 2004
@@ -225,7 +225,7 @@ Description-Content-Type: text/markdown
 
 **AI-ready change intelligence for Java/Spring enterprise monoliths.**
 
-![Version](https://img.shields.io/badge/version-1.31.30-blue)
+![Version](https://img.shields.io/badge/version-1.31.31-blue)
 ![Python](https://img.shields.io/badge/python-3.10%2B-green)
 
 ---
@@ -263,7 +263,7 @@ pipx install sourcecode
 
 ```bash
 sourcecode version
-# sourcecode 1.31.30
+# sourcecode 1.31.31
 ```
 
 ---

{sourcecode-1.31.30 → sourcecode-1.31.31}/README.md

@@ -2,7 +2,7 @@
 
 **AI-ready change intelligence for Java/Spring enterprise monoliths.**
 
-![Version](https://img.shields.io/badge/version-1.31.30-blue)
+![Version](https://img.shields.io/badge/version-1.31.31-blue)
 ![Python](https://img.shields.io/badge/python-3.10%2B-green)
 
 ---
@@ -40,7 +40,7 @@ pipx install sourcecode
 
 ```bash
 sourcecode version
-# sourcecode 1.31.30
+# sourcecode 1.31.31
 ```
 
 ---

{sourcecode-1.31.30 → sourcecode-1.31.31}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "sourcecode"
-version = "1.31.30"
+version = "1.31.31"
 description = "Deterministic codebase context for AI coding agents"
 readme = "README.md"
 requires-python = ">=3.9"

{sourcecode-1.31.30 → sourcecode-1.31.31}/src/sourcecode/__init__.py

@@ -1,3 +1,3 @@
 """sourcecode — Deterministic codebase context maps for AI coding agents."""
 
-__version__ = "1.31.30"
+__version__ = "1.31.31"

{sourcecode-1.31.30 → sourcecode-1.31.31}/src/sourcecode/cli.py

@@ -171,6 +171,8 @@ _SUBCOMMANDS: frozenset[str] = frozenset(
         "repo-ir", "mcp", "endpoints", "impact",
         # Enterprise workflow commands
         "onboard", "modernize", "fix-bug", "review-pr",
+        # License
+        "activate",
     }
 )
 
@@ -253,6 +255,35 @@ def _emit_error_json(error: str, message: str, **context: object) -> None:
     _sys.stderr.flush()
 
 
+# H-06: Intercept Click-level UsageError (unknown options, bad args) and emit JSON.
+# Click's default show() writes "Error: No such option: --foo" as plain text.
+# Automation consumers need JSON on stderr regardless of how the error originated.
+try:
+    import click.exceptions as _click_exc
+
+    def _json_click_usage_error_show(self: Any, file: Any = None) -> None:  # type: ignore[override]
+        import json as _je
+        import sys as _jse
+        _code_map = {
+            "NoSuchOption": "invalid_option",
+            "BadOptionUsage": "invalid_option",
+            "BadParameter": "bad_parameter",
+            "MissingParameter": "missing_required",
+            "BadArgumentUsage": "bad_argument",
+        }
+        code = _code_map.get(type(self).__name__, "invalid_option")
+        payload: dict[str, object] = {"error": code, "message": self.format_message()}
+        _opt = getattr(self, "option_name", None) or getattr(self, "param_hint", None)
+        if _opt:
+            payload["flag"] = str(_opt).strip("'\"")
+        _jse.stderr.write(_je.dumps(payload, ensure_ascii=False) + "\n")
+        _jse.stderr.flush()
+
+    _click_exc.UsageError.show = _json_click_usage_error_show  # type: ignore[method-assign]
+except Exception:
+    pass  # click unavailable — plain-text fallback
+
+
 def _copy_to_clipboard(content: str) -> bool:
     """Copy text to system clipboard. Returns True on success, False otherwise (never raises)."""
     import subprocess
@@ -2165,6 +2196,12 @@ def prepare_context_cmd(
         )
         raise typer.Exit(code=1)
 
+    # Pro gate: generate-tests and delta require an active Pro license.
+    _PRO_TASKS: frozenset[str] = frozenset({"generate-tests", "delta"})
+    if task in _PRO_TASKS:
+        from sourcecode.license import require_pro as _require_pro
+        _require_pro(task)
+
     # Validate --format: only "json" and "github-comment" are valid for prepare-context.
     # "yaml" is intentionally NOT supported here (use main command for yaml output).
     # Invalid values must error loudly — silently falling through to JSON is a lie.
@@ -2222,7 +2259,49 @@ def prepare_context_cmd(
             _sys.stderr.flush()
     _t0 = _time.perf_counter()
     try:
-        output = builder.build(task, since=since, symptom=symptom, fast=fast, include_config=include_config, all_gaps=all_gaps)
+        # H-02: apply timeout for generate-tests — large repos can stall indefinitely.
+        # Mirrors SOURCECODE_TESTS_TIMEOUT_MS used by the MCP generate_tests_context tool.
+        if task == "generate-tests" and not fast:
+            import concurrent.futures as _cf
+            import os as _os_gt
+            _timeout_ms = int(_os_gt.environ.get("SOURCECODE_TESTS_TIMEOUT_MS", "30000"))
+            _timeout_s = _timeout_ms / 1000.0
+            _ex = _cf.ThreadPoolExecutor(max_workers=1)
+            _fut = _ex.submit(
+                builder.build, task,
+                since=since, symptom=symptom, fast=fast,
+                include_config=include_config, all_gaps=all_gaps,
+            )
+            _done_set, _nd_set = _cf.wait([_fut], timeout=_timeout_s)
+            _ex.shutdown(wait=False)
+            if _nd_set:
+                import sys as _sys_gt
+                if _sys_gt.stderr.isatty():
+                    _sys_gt.stderr.write(
+                        f"[generate-tests] timeout after {_timeout_ms}ms — returning partial result\n"
+                    )
+                    _sys_gt.stderr.flush()
+                from sourcecode.prepare_context import TaskOutput as _TO
+                output = _TO(
+                    task=task,
+                    goal=TASKS[task].goal,
+                    project_summary=None,
+                    architecture_summary=None,
+                    relevant_files=[],
+                    suspected_areas=[],
+                    improvement_opportunities=[],
+                    test_gaps=[],
+                    key_dependencies=[],
+                    code_notes_summary=None,
+                    limitations=[f"generate-tests timed out after {_timeout_ms}ms"],
+                    truncated=True,
+                    truncated_reason=f"timeout_{_timeout_ms}ms",
+                    confidence="low",
+                )
+            else:
+                output = _fut.result()
+        else:
+            output = builder.build(task, since=since, symptom=symptom, fast=fast, include_config=include_config, all_gaps=all_gaps)
     finally:
         _progress.finish()
     _t_total = (_time.perf_counter() - _t0) * 1000
@@ -2532,6 +2611,19 @@ def prepare_context_cmd(
     if llm_prompt:
         out["llm_prompt"] = builder.render_prompt(output)
 
+    # H-01: fast-mode analysis transparency — consumer must not confuse "not analyzed"
+    # with "analyzed and found nothing". Fields that were never computed are absent or null,
+    # not zero. analysis_mode and skipped_analyzers make the omission explicit.
+    if fast:
+        out["analysis_mode"] = "fast"
+        _skipped: list[str] = ["deep_content_scan"]
+        _spec = TASKS.get(task)
+        if _spec and _spec.enable_code_notes:
+            _skipped.append("code_notes")
+        if task == "generate-tests":
+            _skipped.append("test_gap_discovery")
+        out["skipped_analyzers"] = _skipped
+
     # P0-1: Apply output budget per task — safety net for large repos.
     from sourcecode.output_budget import (
         trim_to_budget as _pc_trim,
@@ -2864,6 +2956,9 @@ def impact_cmd(
       sourcecode impact org.keycloak.services.DefaultKeycloakSession /path/to/keycloak
       sourcecode impact UserService --depth 6 --output impact.json
     """
+    from sourcecode.license import require_pro as _require_pro
+    _require_pro("impact")
+
     import json as _json
     import sys as _sys
 
@@ -2924,9 +3019,11 @@ def impact_cmd(
             if _copy_to_clipboard(output):
                 typer.echo("✓ copied to clipboard", err=True)
 
-    # Non-zero exit when target not found
+    # H-03: resolution=not_found is a valid structured answer, not an infra failure.
+    # Exit 0 so pipelines can parse the JSON without treating it as an error.
+    # Exit 1 is reserved for path-not-found, I/O failures, and real infra errors.
     if result.get("resolution") == "not_found":
-        raise typer.Exit(code=1)
+        raise typer.Exit(code=0)
 
     from sourcecode.mcp_nudge import nudge_mcp_if_needed as _nudge
     _nudge()
@@ -3246,6 +3343,9 @@ def modernize_cmd(
       sourcecode onboard .         — Architecture overview first
       sourcecode impact <target>   — Verify impact before touching a hotspot
     """
+    from sourcecode.license import require_pro as _require_pro
+    _require_pro("modernize")
+
     import json as _json
     import sys as _sys
     from sourcecode.repository_ir import build_repo_ir, find_java_files, apply_ir_size_limits
@@ -3418,6 +3518,24 @@ def modernize_cmd(
 
 # ── version ───────────────────────────────────────────────────────────────────
 
+@app.command("activate")
+def activate_cmd(
+    license_key: str = typer.Argument(..., help="Your Pro license key"),
+) -> None:
+    """Activate a Pro license key.
+
+    \b
+    Validates the key against the license server and writes
+    ~/.sourcecode/license.json.
+
+    \b
+    Examples:
+      sourcecode activate SC-XXXX-XXXX-XXXX
+    """
+    from sourcecode.license import activate_license as _activate
+    _activate(license_key)
+
+
 @app.command("version")
 def version_cmd() -> None:
     """Show version and exit."""
@@ -3471,6 +3589,9 @@ def mcp_serve() -> None:
         }
       }
     """
+    from sourcecode.license import require_pro as _require_pro
+    _require_pro("mcp serve")
+
     import logging
     import sys as _sys
 

sourcecode-1.31.31/src/sourcecode/license.py

@@ -0,0 +1,166 @@
+"""License activation and enforcement for the sourcecode CLI.
+
+Flow:
+  1. Module imported → _init() loads ~/.sourcecode/license.json (if present)
+  2. is_pro set globally (True when plan == "pro")
+  3. Pro commands call require_pro(feature_name) at entry — exits 1 if not Pro
+  4. `sourcecode activate <key>` calls activate_license(key) — validates via
+     Supabase REST, writes ~/.sourcecode/license.json, exits 0 on success
+
+Supabase credentials:
+  SOURCECODE_SUPABASE_URL      — project REST endpoint
+  SOURCECODE_SUPABASE_ANON_KEY — public anon key (not a secret)
+"""
+from __future__ import annotations
+
+import json
+import os
+import sys
+from datetime import datetime, timezone
+from pathlib import Path
+from typing import Optional
+
+# ---------------------------------------------------------------------------
+# Supabase endpoint config — override via env vars
+# ---------------------------------------------------------------------------
+_SUPABASE_URL: str = os.environ.get(
+    "SOURCECODE_SUPABASE_URL",
+    "https://YOUR_PROJECT.supabase.co",
+)
+_SUPABASE_ANON_KEY: str = os.environ.get(
+    "SOURCECODE_SUPABASE_ANON_KEY",
+    "",
+)
+
+_LICENSE_DIR: Path = Path.home() / ".sourcecode"
+_LICENSE_FILE: Path = _LICENSE_DIR / "license.json"
+
+# ---------------------------------------------------------------------------
+# Global license state — loaded once at import time
+# ---------------------------------------------------------------------------
+_license_data: Optional[dict] = None
+is_pro: bool = False
+
+
+def _load_license_file() -> Optional[dict]:
+    """Read ~/.sourcecode/license.json. Returns parsed dict or None."""
+    try:
+        if _LICENSE_FILE.exists():
+            raw = _LICENSE_FILE.read_text(encoding="utf-8")
+            return json.loads(raw)
+    except Exception:
+        pass
+    return None
+
+
+def _init() -> None:
+    global _license_data, is_pro
+    _license_data = _load_license_file()
+    is_pro = (
+        _license_data is not None
+        and _license_data.get("plan") == "pro"
+    )
+
+
+_init()
+
+
+# ---------------------------------------------------------------------------
+# Enforcement
+# ---------------------------------------------------------------------------
+
+def require_pro(feature_name: str) -> None:
+    """Exit with structured JSON error when not Pro.
+
+    Call at the very top of every Pro-gated command, before any work.
+
+    Example:
+        from sourcecode.license import require_pro
+        require_pro("impact")
+    """
+    if not is_pro:
+        payload = {
+            "error": "pro_required",
+            "feature": feature_name,
+            "message": "Run sourcecode activate <license_key>",
+        }
+        sys.stdout.write(json.dumps(payload, ensure_ascii=False) + "\n")
+        sys.stdout.flush()
+        sys.exit(1)
+
+
+# ---------------------------------------------------------------------------
+# Activation
+# ---------------------------------------------------------------------------
+
+def activate_license(license_key: str) -> None:
+    """Validate license_key via Supabase, write ~/.sourcecode/license.json.
+
+    Outputs JSON to stdout; exits 0 on success, 1 on any failure.
+    Never raises — all error paths emit JSON and call sys.exit(1).
+    """
+    import urllib.error
+    import urllib.request
+
+    # Bail early when Supabase isn't configured yet
+    if not _SUPABASE_ANON_KEY or _SUPABASE_URL == "https://YOUR_PROJECT.supabase.co":
+        _fail("configuration_error", "SOURCECODE_SUPABASE_URL / SOURCECODE_SUPABASE_ANON_KEY not configured.")
+
+    url = (
+        f"{_SUPABASE_URL}/rest/v1/users"
+        f"?license_key=eq.{license_key}"
+        f"&select=license_key,plan,email"
+    )
+    req = urllib.request.Request(url)
+    req.add_header("apikey", _SUPABASE_ANON_KEY)
+    req.add_header("Authorization", f"Bearer {_SUPABASE_ANON_KEY}")
+    req.add_header("Accept", "application/json")
+
+    try:
+        with urllib.request.urlopen(req, timeout=10) as resp:
+            body = resp.read().decode("utf-8")
+    except urllib.error.HTTPError as exc:
+        _fail("network_error", f"Supabase returned HTTP {exc.code}")
+    except Exception as exc:
+        _fail("network_error", str(exc))
+
+    try:
+        rows = json.loads(body)
+    except Exception:
+        _fail("network_error", "Invalid JSON response from Supabase")
+
+    if not rows:
+        _fail("invalid_license", "License key not found")
+
+    user = rows[0]
+    if user.get("plan") != "pro":
+        _fail("not_pro", "This license is not Pro")
+
+    # Write license file
+    _LICENSE_DIR.mkdir(parents=True, exist_ok=True)
+    data = {
+        "license_key": license_key,
+        "plan": "pro",
+        "email": user.get("email", ""),
+        "activated_at": datetime.now(timezone.utc).isoformat(),
+    }
+    _LICENSE_FILE.write_text(
+        json.dumps(data, indent=2, ensure_ascii=False),
+        encoding="utf-8",
+    )
+
+    result = {"status": "activated", "plan": "pro"}
+    sys.stdout.write(json.dumps(result, ensure_ascii=False) + "\n")
+    sys.stdout.flush()
+
+
+# ---------------------------------------------------------------------------
+# Internal helper
+# ---------------------------------------------------------------------------
+
+def _fail(error: str, message: str) -> None:
+    """Emit JSON error to stdout and exit 1. Never returns."""
+    payload = {"error": error, "message": message}
+    sys.stdout.write(json.dumps(payload, ensure_ascii=False) + "\n")
+    sys.stdout.flush()
+    sys.exit(1)

{sourcecode-1.31.30 → sourcecode-1.31.31}/src/sourcecode/mcp/server.py

@@ -88,6 +88,27 @@ def _normalize_repo_path(path: str) -> str:
     return path
 
 
+def _check_repo_path(path: str) -> "CallToolResult | None":
+    """H-05: Validate repo_path exists and is a directory before executing.
+
+    Returns a structured CallToolResult(isError=True) when the path is invalid,
+    or None when the path is valid. Must be called after _normalize_repo_path().
+    Early validation prevents the MCP server from hanging when the CLI exits
+    non-zero with an empty stdout (error went to stderr, not captured by runner).
+    """
+    if not os.path.exists(path):
+        return _err(
+            f"directory_not_found: '{path}' does not exist.",
+            "DIRECTORY_NOT_FOUND",
+        )
+    if not os.path.isdir(path):
+        return _err(
+            f"not_a_directory: '{path}' is not a directory.",
+            "NOT_A_DIRECTORY",
+        )
+    return None
+
+
 @mcp.tool()
 def get_compact_context(repo_path: str = ".", git_context: bool = False) -> dict:
     """Compact human/LLM summary of a repository (~1000-3000 tokens). USE THIS FIRST.
@@ -108,6 +129,9 @@ def get_compact_context(repo_path: str = ".", git_context: bool = False) -> dict
         if not isinstance(git_context, bool):
             return _err("git_context must be boolean", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         args = [repo_path, "--compact"]
         if git_context:
             args.append("--git-context")
@@ -139,6 +163,9 @@ def get_agent_context(repo_path: str = ".", git_context: bool = False) -> dict:
         if not isinstance(git_context, bool):
             return _err("git_context must be boolean", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         args = [repo_path, "--agent"]
         if git_context:
             args.append("--git-context")
@@ -174,6 +201,9 @@ def get_endpoints(repo_path: str = ".") -> dict:
         if not isinstance(repo_path, str):
             return _err("repo_path must be a string", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         return _execute(["endpoints", repo_path])
     except Exception as exc:
         return _err(
@@ -197,6 +227,9 @@ def get_module_context(repo_path: str = ".", module: str = "") -> dict:
         if not isinstance(module, str) or not module.strip():
             return _err("module must be a non-empty string", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         module_path = repo_path.rstrip("/") + "/" + module.strip("/")
         return _execute([module_path, "--compact"])
     except Exception as exc:
@@ -221,6 +254,9 @@ def get_delta(repo_path: str = ".", since: str = "HEAD~1") -> dict:
         if not isinstance(since, str) or not since.strip():
             return _err("since must be a non-empty git ref", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         return _execute(["prepare-context", "delta", repo_path, "--since", since])
     except Exception as exc:
         return _err(
@@ -248,6 +284,9 @@ def get_ir_summary(repo_path: str = ".") -> dict:
         if not isinstance(repo_path, str):
             return _err("repo_path must be a string", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         return _execute(["repo-ir", repo_path, "--summary-only"])
     except Exception as exc:
         return _err(
@@ -271,6 +310,9 @@ def fix_bug_context(repo_path: str = ".", symptom: str = "") -> dict:
         if not isinstance(repo_path, str):
             return _err("repo_path must be a string", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         args = ["prepare-context", "fix-bug", repo_path]
         if symptom and isinstance(symptom, str) and symptom.strip():
             args.extend(["--symptom", symptom.strip()])
@@ -297,6 +339,9 @@ def review_pr_context(repo_path: str = ".", since: str = "") -> dict:
         if not isinstance(repo_path, str):
             return _err("repo_path must be a string", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         args = ["prepare-context", "review-pr", repo_path]
         if since and isinstance(since, str) and since.strip():
             args.extend(["--since", since.strip()])
@@ -320,6 +365,9 @@ def onboard_context(repo_path: str = ".") -> dict:
         if not isinstance(repo_path, str):
             return _err("repo_path must be a string", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         return _execute(["prepare-context", "onboard", repo_path])
     except Exception as exc:
         return _err(
@@ -341,6 +389,9 @@ def explain_context(repo_path: str = ".") -> dict:
         if not isinstance(repo_path, str):
             return _err("repo_path must be a string", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         return _execute(["prepare-context", "explain", repo_path])
     except Exception as exc:
         return _err(
@@ -362,6 +413,9 @@ def refactor_context(repo_path: str = ".") -> dict:
         if not isinstance(repo_path, str):
             return _err("repo_path must be a string", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         return _execute(["prepare-context", "refactor", repo_path])
     except Exception as exc:
         return _err(
@@ -388,6 +442,9 @@ def generate_tests_context(repo_path: str = ".", include_all: bool = False) -> d
         if not isinstance(include_all, bool):
             return _err("include_all must be boolean", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         args = ["prepare-context", "generate-tests", repo_path]
         if include_all:
             args.append("--all")
@@ -444,6 +501,9 @@ def get_impact_context(repo_path: str = ".", target: str = "", depth: int = 4) -
         if not isinstance(depth, int) or depth < 1 or depth > 8:
             return _err("depth must be an integer between 1 and 8", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         args = ["impact", target.strip(), repo_path, "--depth", str(depth)]
         return _execute(args)
     except Exception as exc:
@@ -475,6 +535,9 @@ def modernize_context(repo_path: str = ".", format: str = "json") -> dict:
         if not isinstance(format, str) or format not in ("json", "yaml"):
             return _err("format must be 'json' or 'yaml'", "INVALID_ARGUMENT")
         repo_path = _normalize_repo_path(repo_path)
+        _path_err = _check_repo_path(repo_path)
+        if _path_err is not None:
+            return _path_err
         return _execute(["modernize", repo_path])
     except Exception as exc:
         return _err(

{sourcecode-1.31.30 → sourcecode-1.31.31}/src/sourcecode/repository_ir.py

@@ -1524,6 +1524,31 @@ def _get_git_old_content(git_root: Path, rel_path: str, since: str) -> Optional[
     return None
 
 
+def _get_git_changed_files(root: "Path", since: str) -> "Optional[frozenset[str]]":
+    """H-04: Return set of paths changed between `since` and HEAD, or None on failure.
+
+    One `git diff --name-only` call replaces O(n) `git show` calls — only files
+    in the returned set need old-content fetched for symbol diff computation.
+    Returns None when git is unavailable or the ref cannot be resolved; the
+    caller must fall back to the original per-file fetch in that case.
+    """
+    try:
+        result = subprocess.run(
+            ["git", "diff", "--name-only", since, "HEAD"],
+            cwd=str(root),
+            capture_output=True,
+            text=True,
+            encoding="utf-8",
+            errors="replace",
+            timeout=10,
+        )
+        if result.returncode == 0:
+            return frozenset(p.strip() for p in result.stdout.splitlines() if p.strip())
+    except (subprocess.TimeoutExpired, OSError, FileNotFoundError):
+        pass
+    return None
+
+
 # ---------------------------------------------------------------------------
 # Phase 5 — Evidence Engine
 # ---------------------------------------------------------------------------
@@ -2511,6 +2536,12 @@ def build_repo_ir(
     all_changed: list[ChangedSymbol] = []
     all_route_diffs: list[dict] = []
 
+    # H-04: prefetch changed-file list once; avoids O(n) `git show` calls.
+    # _since_changed=None means git unavailable → fall back to per-file fetch.
+    _since_changed: "Optional[frozenset[str]]" = None
+    if since:
+        _since_changed = _get_git_changed_files(root, since)
+
     for rel_path in sorted(file_paths):
         abs_path = root / rel_path
         try:
@@ -2520,7 +2551,11 @@ def build_repo_ir(
 
         old_source: Optional[str] = None
         if since:
-            old_source = _get_git_old_content(root, rel_path, since)
+            # Only fetch old content for files known to have changed.
+            # Unchanged files have no diff entries — skip git show entirely.
+            _file_changed = _since_changed is None or rel_path in _since_changed
+            if _file_changed:
+                old_source = _get_git_old_content(root, rel_path, since)
 
         package, symbols, raw_imports = _extract_symbols(source, rel_path)
         relations = _build_relations(symbols, raw_imports, source, package, rel_path)
@@ -2529,7 +2564,8 @@ def build_repo_ir(
             _, old_symbols, _ = _extract_symbols(old_source, rel_path)
             all_changed.extend(_diff_symbols(old_symbols, symbols))
             all_route_diffs.extend(_diff_routes(old_symbols, symbols))
-        elif since:
+        elif since and (_since_changed is None or rel_path in _since_changed):
+            # File is new in since..HEAD (not in old ref) — treat as added.
             for sym in symbols:
                 all_changed.append(ChangedSymbol(
                     symbol=sym.symbol,