Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →The claim that Claude Code’s Read deny rules let four of eleven routes through should be treated as an unverified test report, not a general product behavior. The available material does not identify the tested version, deny pattern, eleven routes, or which four succeeded. What the documentation does establish is narrower: deny rules are evaluated before permission modes, while native file reads, shell commands, and CLAUDE.md imports are distinct mechanisms that need separate evaluation.
What the “four of eleven” claim does—and does not—show
The headline’s count is not independently substantiated by the available sources. They do not provide an experiment log or enough configuration detail to reproduce the result. In particular, they do not establish whether grep -r, a Python file-reading one-liner, or either of the two CLAUDE.md @imports were among the routes that succeeded.
As an Amazon Associate I earn from qualifying purchases.
A third-party PyPI project description for deny-probe lists examples involving grep -r, a Python file-reading one-liner, and CLAUDE.md import chains. That description is not evidence that the project produced the four-of-eleven result, or that those examples bypassed a particular Claude Code rule. The count and the identity of any successful routes remain unverified.
Free tools Windows power users keep installed
One-click scans. No signup required.
What Claude Code’s deny rules are documented to do
Claude Code’s SDK documentation says deny rules are evaluated before permission modes. A matching deny blocks the call even when bypassPermissions is selected. That describes how a matching rule is handled; it does not mean that every way of accessing file contents necessarily matches the same rule.
A tool-name deny and a scoped pattern are different controls
A bare tool-name deny removes that tool. A scoped pattern instead blocks calls matching its pattern. Their coverage therefore differs: the first targets use of a tool, while the second depends on the calls that actually match the specified pattern. Without the exact rule and route, a test result cannot establish broader coverage.
Permission mode and version still matter to route analysis
Claude Code’s permission-mode documentation describes recognized file-reading Bash commands and special handling for reads outside working directories when the relevant setting is enabled. The documented outside-read behavior requires Claude Code v2.1.257 or later. This version-specific behavior is a reason to record the installed version and settings rather than infer that a route will behave identically across releases.
Why Read, Bash, and CLAUDE.md imports should be tested separately
Native file reads
A native Read tool call is a direct test of a rule aimed at that tool or its matching calls. Record the exact tool call and whether the protected file’s contents were returned; an attempted call and a successful read are not the same outcome.
Shell readers such as grep -r
A shell command that reads files is a different path from a native Read call. Claude Code documentation describes recognition of certain file-reading Bash commands, but the retrieved material does not establish that every shell reader—or an arbitrary subprocess—will be governed in the same way. Test the exact command, its working directory, relevant settings, and whether the output actually exposes protected content.
Rank #3
Python one-liners and indirect scripts
A Python one-liner that opens a file is not interchangeable with a recognized shell reader or a native Read call. The available sources do not establish how a specific one-liner is treated under a specific configuration. Test the exact invocation rather than generalizing from the presence of a deny rule.
CLAUDE.md @path imports
Claude Code supports @path imports in CLAUDE.md, and its documentation describes how project memory files are loaded. That fact alone does not show that an import bypasses a permission rule. Keep two questions separate: whether an instruction file was loaded, and whether the contents of the protected target file were read.
How to reproduce a deny-rule test responsibly
To assess a specific configuration, use a disposable protected file with a harmless marker and record the complete setup. A useful report includes:
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- Claude Code version and the operating environment.
- The exact deny rule and any relevant permission settings, including the selected permission mode.
- The protected file, its location relative to the working directory, and the setting relevant to outside reads.
- Every tested route, including the exact native tool call, shell command, script, or import path.
- For each attempt, the expected result, observed result, and whether a prompt or human approval occurred.
- Whether the protected file’s contents or marker were actually exposed, rather than merely whether the action was attempted.
Report each route as its own result. A blocked native Read call does not establish that a shell command or import was blocked, and a loaded instruction file does not establish that a protected target was read. Repeating the test after a version or configuration change is necessary before extending the result to that new setup.
Best Value
How to evaluate the protection you have configured
Check the control against the access path you intend to constrain. Determine whether it targets a tool name or a scoped command pattern, then test native file tools, documented Bash readers, and indirect script routes separately. Include the applicable permission mode and version in the assessment.
Hooks and advisory instructions in CLAUDE.md are distinct from permission rules. The available sources establish that these mechanisms exist, but they do not certify any one configuration as complete against every route in the reported test. Anthropic’s auto-mode engineering article says auto mode drops permission rules known to grant arbitrary code execution, including blanket shell access and wildcarded script interpreters. That is about auto mode’s handling of risky allow rules; it is not a universal guarantee of Read-deny coverage.
What can be concluded
The defensible conclusion is that a matching deny rule is documented to block its matching call even in bypassPermissions mode, but the rule’s scope and each access mechanism matter. The specific four-of-eleven result—and which routes, if any, account for it—cannot be confirmed from the available evidence.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
