Back to skill

Security audit

the compliance claw

Security checks for vulnerabilities and agentic risk

Overview

The skill’s stated compliance-monitoring purpose is coherent, but it installs an unpinned third-party Homebrew executable that would handle sensitive client compliance data.

Review the Homebrew tap, publisher, release version, and source before installing. Use test or non-production compliance records first, and avoid granting calendar or sensitive client-data access until the CLI’s data handling and update provenance are clear.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (1)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:6
Finding
Unpinned Executable Installed from a Third-Party Homebrew Tap## Vulnerability Details **File Location**: `SKILL.md`, line 6 **Vulnerability Type**: Unpinned third-party dependency from a mutable external source **Risk Level**: Medium ### Vulnerable Code ```yaml metadata: {"clawdbot":{"emoji":"🏛️","requires":{"bins":["complianceclaw"]},"install":[{"id":"brew","kind":"brew","formula":"legal-tools/tap/complianceclaw","bins":["complianceclaw"],"label":"Install complianceclaw (brew)"}]}} ``` ### Technical Analysis The skill declares `legal-tools/tap/complianceclaw`, a package hosted through a third-party custom Homebrew tap, as the installation source for the required executable. The declaration does not pin an immutable package version or source revision and does not specify an expected cryptographic checksum. Consequently, the code installed by this instruction can change after the skill has been reviewed. The repository contains only documentation and metadata, so the implementation, installation logic, and runtime behavior of the external `complianceclaw` executable cannot be audited from this package. This creates a supply-chain trust boundary: anyone who compromises or gains publishing control over the custom tap, its formula, or its referenced release artifacts could substitute malicious code. This finding establishes an insecure dependency risk; it does not, by itself, establish that the current external package is malicious. ### Attack Path 1. An attacker compromises the custom Homebrew tap, its publisher account, or an upstream artifact referenced by its formula. 2. The attacker changes the formula or release artifact while retaining the expected package and binary names. 3. A user or agent follows the skill metadata and installs `legal-tools/tap/complianceclaw`. 4. Homebrew retrieves and installs the attacker-controlled package because the skill does not constrain installation to a reviewed immutable version. 5. Malicious installation logic or the substituted `compliancecla ...[truncated 1109 chars]
Remediation
## Remediation Suggestions 1. Pin the dependency to an immutable, reviewed release rather than a mutable formula reference. 2. Verify downloaded artifacts using a cryptographic checksum such as SHA-256 and ensure the expected digest is independently maintained. 3. Pin the Homebrew tap or formula to a reviewed commit where the installation framework supports doing so. 4. Publish the CLI source and build instructions, and provide reproducible builds so users can verify that release artifacts correspond to reviewed source. 5. Use signed releases and document how consumers should verify the publisher signature and signing identity. 6. Document the exact official repository, package publisher, version, and expected artifact digest in the skill. 7. Review the formula for installation hooks or commands that execute during installation, not only the final binary. 8. Run the CLI with least privilege and isolate it from unrelated credentials and sensitive files. 9. Avoid supplying production client records until the executable's implementation, data handling, and outbound network behavior have been independently assessed.
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Static analysis

No suspicious patterns detected.