Back to skill

Security audit

Apple Calendar CLI

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward Apple Calendar helper, but it grants read/write/delete access to sensitive calendar data without enough safeguards and installs an unpinned third-party CLI.

Review this skill before installing. Only grant Calendar access if you trust the third-party CLI source, prefer a pinned and verified release if available, and require the agent to ask before reading broad date ranges or changing/deleting events.

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:8
Finding
Unpinned Third-Party Homebrew Tap Creates a Supply-Chain Risk## Vulnerability Details **File Location**: `SKILL.md`, line 8 **Vulnerability Type**: Unpinned third-party dependency from a custom Homebrew tap **Risk Level**: Medium ### Vulnerable Code Snippet ```markdown - Install: `brew install sichengchen/tap/apple-calendar-cli` ``` ### Technical Analysis The skill directs users or agents to install `apple-calendar-cli` from a third-party custom Homebrew tap without pinning a reviewed version, formula revision, source commit, checksum, or artifact signature. Because the dependency is mutable and its implementation is not included in this project, the reviewed skill text cannot guarantee that the executable installed later will be the same executable that was previously assessed. Homebrew installation may execute formula-defined installation logic and place an executable on the user's command path. This trust boundary is particularly sensitive because the documented tool is expected to receive macOS Calendar permission. Once authorized, it can enumerate calendars, read event details, and create, update, move, or delete events. The audit did not establish that the referenced package is malicious; the issue is the absence of dependency pinning and integrity verification. ### Attack Path 1. An attacker compromises the third-party Homebrew tap, its maintainer account, or an upstream release location referenced by the formula. 2. The attacker modifies the formula or replaces an expected release artifact with a malicious version. 3. A user or agent follows the instruction: ```bash brew install sichengchen/tap/apple-calendar-cli ``` 4. Homebrew retrieves and installs the attacker-controlled dependency or executes compromised formula logic. 5. The installed program executes with the privileges of the invoking user. 6. If Calendar access is granted as instructed, the malicious program can access or alter calendar data available through that permission. It may also access other resources already available to the in ...[truncated 736 chars]
Remediation
## Remediation Suggestions 1. Pin the dependency to a reviewed, immutable version and document that exact version in the skill. 2. Pin or otherwise verify the Homebrew formula revision rather than implicitly trusting the latest state of a mutable tap. 3. Verify downloaded artifacts using a published cryptographic checksum or trusted signature. Store the expected digest in reviewed project material. 4. Link to the canonical source repository, release page, and security policy so reviewers can validate package provenance. 5. Prefer an official package source where available. If a custom tap remains necessary, document its ownership and trust assumptions. 6. Vendor the relevant auditable source or installation definition when feasible so changes are visible during skill review. 7. Apply least privilege: grant Calendar access only to the intended signed executable, review macOS authorization prompts, and revoke access when the tool is no longer required. 8. In automated environments, validate the resolved package version and checksum before executing the installed binary, and fail closed on any mismatch.
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
Findings (2)

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The skill enables listing calendars and events and retrieving full event details, which can expose sensitive personal data such as meeting titles, locations, notes, URLs, attendees, and schedule patterns. Because the skill tells agents to use these commands directly and lacks privacy warnings or data-minimization guidance, it creates a meaningful risk of over-collection and disclosure of personal scheduling information.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill documents deletion and modification workflows for calendar events without any guidance to obtain explicit user confirmation before destructive actions. In an agent context, this increases the risk of accidental or prompt-induced deletion/rescheduling of real user data, since calendar operations affect personal or business schedules and may be irreversible in practice.

Static analysis

No suspicious patterns detected.