📦

Audit History

socialdatax-media-save-organizer - 4 audits

Version comparison

Capability and finding changes across audited versions, newest first.

VersionDateResultReview itemsChange vs previous
v4 LatestOct 29, 2026, 10:42 AM No confirmed findings2No capability change
v3 Oct 29, 2026, 04:55 AM No confirmed findings2No capability change
v2 Oct 21, 2026, 09:34 AM 2 confirmed0No capability change
v1 Oct 21, 2026, 08:19 AM 2 confirmed0Baseline

Oct 29, 2026, 10:42 AM

Thirty-six alerts are false positives involving Markdown, Chinese text, service links, or documented credential configuration. Two command locations execute an unpinned npm package without an installation prompt, creating supply chain exposure. No evidence found of malicious intent, credential exfiltration, or prompt injection in the reviewed files; external package behavior was not verified.

2
Files scanned
148
Lines analyzed
5
Review items
0
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
Ruby/shell backtick execution
```bash
The fenced workflow executes npx -y socialdatax-skills@latest, resolving a mutable package without an installation prompt. Compromised releases could execute with local permissions and credentials.
Medium
Ruby/shell backtick execution
将以下本地保存子命令接在 `npx -y socialdatax-skills@latest` 后执行;纯本地 `download-media` 不接受 `--source-client`、`--so
This instruction explicitly executes download subcommands through npx -y socialdatax-skills@latest. Unpinned package resolution can run changed third-party code with local user permissions.
Audited by: codex

Oct 29, 2026, 04:55 AM

Most detections are Markdown backticks, normal API authentication references, documentation links, or readable Chinese text rather than shell substitution, leaked secrets, or obfuscation. Two command locations execute an unpinned npm package through npx, exposing the host and inherited credentials to changing third-party code. No evidence of prompt injection or intentional exfiltration was found; external package implementation was not included for review.

2
Files scanned
148
Lines analyzed
5
Review items
0
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
Ruby/shell backtick execution
```bash
Although the delimiter is Markdown, the enclosed commands run npx -y socialdatax-skills@latest. Changing third-party code executes with host permissions and potentially inherited API credentials.
Medium
Ruby/shell backtick execution
将以下本地保存子命令接在 `npx -y socialdatax-skills@latest` 后执行;纯本地 `download-media` 不接受 `--source-client`、`--so
The prose explicitly directs execution through npx -y socialdatax-skills@latest. Automatic execution of an unpinned package permits compromised future releases to access the host.
Audited by: codex

Oct 21, 2026, 09:34 AM

All 37 static alerts match Markdown syntax, documented service configuration, or readable multilingual text rather than the alleged threats. Contextual risks remain in automatic execution of an unpinned npm package and forwarding unvalidated payment links from API errors. No evidence found of prompt injection or credential exfiltration in the supplied files; the external CLI implementation was not provided.

2
Files scanned
148
Lines analyzed
5
Review items
0
False positives ignored

Confirmed security concerns (2)

Medium
Automatic Execution of an Unpinned npm Dependency
Examples use npx -y socialdatax-skills@latest, allowing mutable registry code to execute with local permissions. A compromised release could access files and the configured service key.
The commands explicitly combine @latest with automatic npm confirmation. The risk is mutable dependency execution, not evidence that the current package is malicious.
Medium
Unvalidated Payment Links Promoted from API Errors
Balance handling instructs agents to display error-provided recharge links unchanged and prompt payment. A compromised response could promote a phishing destination without explicit validation.
The text explicitly promotes external error links without a billing-domain check. No malicious destination is shown; exploitation would require an attacker-controlled response.
Audited by: codex

Oct 21, 2026, 08:19 AM

All 38 static alerts are false positives for their reported patterns: Markdown backticks, documented credential names, provider links, and readable Chinese text. Separate semantic risks involve executing an unpinned npm package and requiring token-bearing XHS URLs in shared outputs without redaction. No evidence found of credential harvesting, encoded payloads, or instructions to override this audit.

2
Files scanned
148
Lines analyzed
5
Review items
0
False positives ignored

Confirmed security concerns (2)

High
Mandatory Disclosure of Token-Bearing Source URLs
The skill requires preserving xsec_token in every XHS URL output, storage location, and forwarded link, explicitly forbidding redaction. This can expose access-related link tokens beyond the intended download workflow; the token's exact privileges are not established.
Line 103 explicitly includes forwarding and forbids redacting xsec_token. The disclosure instruction is clear, although the token's scope and lifetime are not documented.
Medium
Unpinned Remote Package Execution
The workflow repeatedly runs npx -y socialdatax-skills@latest. This executes mutable registry code with local permissions and possible access to the configured API key.
The commands explicitly use @latest and suppress the npx installation prompt. No evidence found of a pinned version or integrity verification in this skill.
Audited by: codex