fix(analyzers): broaden and redact secret detection

CHANGELOG.md

@@ -7,6 +7,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+### Security
+
+- Broaden secret detection for AI/provider tokens, modern GitHub tokens, package registry tokens, private keys, and common JSON/YAML assignments; redact matched credentials from finding evidence.
+
 ## [0.1.4] - 2026-06-12
 
 ### Security

src/mcts/analyzers/data_leakage.py

@@ -11,16 +11,33 @@
 
 SECRET_PATTERNS: list[tuple[str, re.Pattern[str], Severity]] = [
     ("OpenAI API Key", re.compile(r"sk-[A-Za-z0-9]{20,}"), Severity.CRITICAL),
+    ("Anthropic API Key", re.compile(r"sk-ant-[A-Za-z0-9_-]{20,}"), Severity.CRITICAL),
+    ("Hugging Face Token", re.compile(r"hf_[A-Za-z0-9]{20,}"), Severity.CRITICAL),
     ("AWS Access Key", re.compile(r"AKIA[0-9A-Z]{16}"), Severity.CRITICAL),
     ("Google API Key", re.compile(r"AIza[0-9A-Za-z\-_]{35}"), Severity.CRITICAL),
     ("Google OAuth Token", re.compile(r"ya29\.[0-9A-Za-z\-_]+"), Severity.CRITICAL),
-    ("GitHub PAT", re.compile(r"ghp_[a-zA-Z0-9]{36}"), Severity.CRITICAL),
+    (
+        "GitHub PAT",
+        re.compile(r"(?:ghp_[a-zA-Z0-9]{36}|github_pat_[a-zA-Z0-9_]{40,})"),
+        Severity.CRITICAL,
+    ),
     ("GitLab PAT", re.compile(r"glpat-[a-zA-Z0-9\-_]{20,}"), Severity.CRITICAL),
     ("Slack Token", re.compile(r"xox[baprs]-[0-9A-Za-z\-]{10,}"), Severity.CRITICAL),
+    ("npm Access Token", re.compile(r"npm_[A-Za-z0-9]{20,}"), Severity.CRITICAL),
+    ("PyPI API Token", re.compile(r"pypi-[A-Za-z0-9_-]{20,}"), Severity.CRITICAL),
+    (
+        "Private Key",
+        re.compile(r"-----BEGIN (?:RSA |EC |OPENSSH )?PRIVATE KEY-----"),
+        Severity.CRITICAL,
+    ),
     ("JWT", re.compile(r"eyJ[a-zA-Z0-9\-_]+\.eyJ[a-zA-Z0-9\-_]+\.[a-zA-Z0-9\-_]+"), Severity.HIGH),
     (
         "Generic Secret Assignment",
-        re.compile(r"(?i)(api_key|secret|password|token)\s*=\s*['\"][^'\"]+['\"]"),
+        re.compile(
+            r"(?i)[\"']?(?:api[_-]?key|client[_-]?secret|access[_-]?token|auth[_-]?token|"
+            r"database[_-]?password|db[_-]?password|password|secret|token)[\"']?\s*[:=]\s*"
+            r"[\"']?[A-Za-z0-9_./+=:@-]{8,}"
+        ),
         Severity.HIGH,
     ),
     ("Database URL", re.compile(r"(?i)(postgres|mysql|mongodb)://\S+"), Severity.HIGH),
@@ -37,6 +54,16 @@
     "DATABASE_URL",
     "GITHUB_TOKEN",
     "ANTHROPIC_API_KEY",
+    "AZURE_OPENAI_API_KEY",
+    "COHERE_API_KEY",
+    "HUGGINGFACE_TOKEN",
+    "LANGFUSE_SECRET_KEY",
+    "LANGSMITH_API_KEY",
+    "MISTRAL_API_KEY",
+    "NPM_TOKEN",
+    "PINECONE_API_KEY",
+    "PYPI_API_TOKEN",
+    "WANDB_API_KEY",
 )
 
 HIDDEN_CHAR_PATTERN = re.compile(r"[\u200b-\u200f\ufeff\u202a-\u202e]")
@@ -60,6 +87,14 @@ def _is_logging_statement(line: str) -> bool:
     return bool(LOGGING_CALL_PATTERN.search(line))
 
 
+def _redact_secrets(line: str) -> str:
+    """Return bounded context with every recognized credential removed."""
+    redacted = line
+    for _, pattern, _ in SECRET_PATTERNS:
+        redacted = pattern.sub("[REDACTED]", redacted)
+    return redacted.strip()[:120]
+
+
 class DataLeakageAnalyzer(BaseAnalyzer):
     """Scans tool metadata and source files for exposed secrets."""
 
@@ -118,7 +153,8 @@ def _scan_source_files(self, server: MCPServerInfo) -> list[Finding]:
         for file_path, content in server.source_files.items():
             for line_no, line in enumerate(content.splitlines(), start=1):
                 for label, pattern, severity in SECRET_PATTERNS:
-                    if not pattern.search(line):
+                    match = pattern.search(line)
+                    if not match:
                         continue
                     if label == "Internal URL" and _is_logging_statement(line):
                         continue
@@ -137,7 +173,7 @@ def _scan_source_files(self, server: MCPServerInfo) -> list[Finding]:
                             technique_id="MCTS-T-1004",
                             confidence=0.7,
                             location=SourceLocation(file=file_path, line=line_no),
-                            evidence={"pattern": pattern.pattern, "line": line.strip()[:120]},
+                            evidence={"pattern": pattern.pattern, "line": _redact_secrets(line)},
                         )
                     )
         return findings

tests/test_analyzers.py

@@ -1,7 +1,10 @@
 """Tests for source-aware analyzers."""
 
+import json
 from pathlib import Path
 
+import pytest
+
 from mcts.analyzers.command_execution import CommandExecutionAnalyzer
 from mcts.analyzers.data_leakage import DataLeakageAnalyzer
 from mcts.core.config import ScanConfig
@@ -50,6 +53,44 @@ def test_data_leakage_ignores_loopback_urls_in_log_messages() -> None:
     assert findings[0].location.line == 4
 
 
+@pytest.mark.parametrize(
+    ("source", "secret"),
+    [
+        ("ANTHROPIC_API_KEY='sk-ant-" + "a" * 30 + "'", "sk-ant-" + "a" * 30),
+        ("token = 'hf_" + "b" * 30 + "'", "hf_" + "b" * 30),
+        ('{"apiKey": "super-secret-value"}', "super-secret-value"),
+        ("DB_PASSWORD: database-password", "database-password"),
+        ("-----BEGIN PRIVATE KEY-----", "-----BEGIN PRIVATE KEY-----"),
+    ],
+)
+def test_data_leakage_detects_and_redacts_common_secret_formats(
+    source: str, secret: str
+) -> None:
+    server = MCPServerInfo(name="secret-fixture", source_files={"config.txt": source})
+
+    findings = DataLeakageAnalyzer().analyze(server)
+
+    assert findings
+    evidence = json.dumps([finding.evidence for finding in findings])
+    assert secret not in evidence
+    assert "[REDACTED]" in evidence
+
+
+def test_data_leakage_redacts_multiple_secrets_from_the_same_line() -> None:
+    anthropic_key = "sk-ant-" + "a" * 30
+    huggingface_token = "hf_" + "b" * 30
+    server = MCPServerInfo(
+        name="two-secrets",
+        source_files={"config.txt": f"primary={anthropic_key} backup={huggingface_token}"},
+    )
+
+    findings = DataLeakageAnalyzer().analyze(server)
+
+    evidence = json.dumps([finding.evidence for finding in findings])
+    assert anthropic_key not in evidence
+    assert huggingface_token not in evidence
+
+
 def test_docker_dedupe_dockerfile_and_containerfile(tmp_path: Path) -> None:
     """Dockerfile + Containerfile with same FROM → only 1 HIGH finding."""
     from mcts.analyzers.supply_chain import SupplyChainAnalyzer