Audit History

Version	Date	Result	Review items	Change vs previous
v5 Latest	Jul 5, 2026, 03:20 PM	No confirmed findings	0	No capability change
v4	Jul 5, 2026, 03:20 PM	No confirmed findings	0	No capability change
v3	Jun 30, 2026, 01:08 AM	No confirmed findings	2	No capability change
v2	Jan 23, 2026, 07:20 AM	No confirmed findings	0	No capability change
v1	Jan 23, 2026, 07:06 AM	No confirmed findings	0	Baseline

Jul 5, 2026, 03:20 PM

The static findings are false positives caused by Markdown inline code and fenced command examples in SKILL.md. The skill documents legitimate Expo and EAS development client workflows, and no prompt injection, exfiltration intent, or hidden automation was found.

1

Files scanned

165

Lines analyzed

1

Review items

0

False positives ignored

Audited by: codex

Audit version 4

Jul 5, 2026, 03:20 PM

The static findings are false positives caused by Markdown inline code and fenced command examples in SKILL.md. The skill documents legitimate Expo and EAS development client workflows, and no prompt injection, exfiltration intent, or hidden automation was found.

1

Files scanned

165

Lines analyzed

1

Review items

0

False positives ignored

Audited by: codex

Audit version 3

Jun 30, 2026, 01:08 AM

Static analysis flagged many backtick and command examples in SKILL.md. Review found these are Markdown instructions for Expo EAS, simulator, device, and Android tooling, not hidden executable code or malicious automation. The skill is publishable with a medium warning because it instructs users to run external commands that can build, install, update tooling, and submit iOS builds.

1

Files scanned

165

Lines analyzed

3

Review items

2

False positives ignored

Capability review items (2)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

Medium

SKILL.md:56-58 SKILL.md:74-80 SKILL.md:90-94 SKILL.md:98-107 SKILL.md:111-130 SKILL.md:141-164

External Build, Install, and Submission Commands

The skill provides shell command examples for EAS cloud builds, local builds, TestFlight submission, simulator installation, device installation, Android installation, credential inspection, cache clearing, and CLI updates. These commands match the skill purpose and are not malicious, but they can change local devices, project state, Expo cloud state, and App Store Connect submissions if run without review.

The referenced lines contain explicit shell commands for known Expo and mobile tooling. The commands are legitimate for this skill, but their operational effects create a real user-consent and environment-change risk.

Low

SKILL.md:78-79 SKILL.md:103 SKILL.md:115-116 SKILL.md:124-130 SKILL.md:160-164

Reconnaissance Findings Are Benign Tooling Checks

The system reconnaissance alerts map to normal build output descriptions and Expo CLI inspection commands. The skill describes checking build status and CLI version, not collecting host data for exfiltration.

The commands and surrounding text are directly tied to normal Expo build workflows. No evidence found of network exfiltration, credential harvesting, or broad host enumeration.

Static false positives ignored (2)

These static matches were dismissed by semantic review or matched schema-only tokens, so they are shown for transparency but do not drive the quality score.

Low

SKILL.md:19 SKILL.md:23-25 SKILL.md:45-50 SKILL.md:56-58 SKILL.md:74-80 SKILL.md:90-94 SKILL.md:105-107 SKILL.md:141-146 SKILL.md:151-164

Markdown Backtick Detection Is Mostly Benign

The static analyzer reported Ruby or shell backtick execution, but the evidence is Markdown inline code and fenced examples in a documentation-only skill. No evidence found of Ruby code, dynamic command construction, hidden execution logic, or automatic command execution.

The file is a Markdown instruction document and the flagged syntax appears inside inline code or fenced command examples. There is no executable wrapper or script file in the scanned structure.

Low

SKILL.md:3 SKILL.md:134

Weak Cryptography Findings Are False Positives

The weak cryptographic algorithm alerts at SKILL.md lines 3 and 134 do not correspond to cryptographic code or algorithm selection. Line 3 is metadata text, and line 134 introduces a feature list for the development client.

The reviewed lines contain ordinary prose about Expo development clients. No evidence found of MD5, SHA-1, DES, RC4, or any other cryptographic implementation.

Detected Patterns

User-Run Shell Commands

Audited by: codex

Audit version 2

Jan 23, 2026, 07:20 AM

All 38 static findings evaluated as false positives. The skill contains legitimate documentation for Expo EAS CLI commands and standard development workflows. No malicious patterns, command injection vectors, or data exfiltration detected.

1

Files scanned

165

Lines analyzed

1

Review items

1

False positives ignored

Static false positives ignored (1)

These static matches were dismissed by semantic review or matched schema-only tokens, so they are shown for transparency but do not drive the quality score.

High

SKILL.md:57 SKILL.md:76 SKILL.md:92 SKILL.md:106

External Commands Documentation

The skill contains examples of EAS CLI and development tool commands. These are legitimate documentation examples for building Expo development clients.

All commands are hardcoded EAS CLI examples with no user input injection. These are standard Expo development workflow documentation.

Audited by: claude

Audit version 1

Jan 23, 2026, 07:06 AM

All 38 static findings are false positives. External command detections are legitimate Expo/EAS CLI commands for development workflows. No malicious patterns, credential exfiltration, or security risks identified. The skill contains only documentation for standard build commands.

1

Files scanned

165

Lines analyzed

1

Review items

0

False positives ignored

Audited by: claude

Version comparison

Risk Factors

Risk Factors

Risk Factors

Detected Patterns

Risk Factors

Risk Factors