Skills video-ad-specs Audit History
📦

Audit History

video-ad-specs - 5 audits

Version comparison

Capability and finding changes across audited versions, newest first.

VersionDateResultReview itemsChange vs previous
v5 LatestJul 9, 2026, 11:43 AM No confirmed findings0No capability change
v4 Jul 9, 2026, 11:43 AM No confirmed findings0No capability change
v3 Jul 5, 2026, 03:15 PM No confirmed findings2No capability change
v2 Jul 5, 2026, 03:15 PM No confirmed findings2No capability change
v1 Jul 3, 2026, 01:13 PM No confirmed findings10Baseline

Jul 9, 2026, 11:43 AM

The static findings are false positives caused by Markdown code fences, inline code, documentation URLs, and advertising terminology. The skill uses transparent belt CLI examples for its stated workflow. No hidden execution, prompt injection, credential access, or malicious exfiltration intent was found.

1
Files scanned
261
Lines analyzed
2
Review items
0
False positives ignored
Audited by: codex

Jul 9, 2026, 11:43 AM

The static findings are false positives caused by Markdown code fences, inline code, documentation URLs, and advertising terminology. The skill uses transparent belt CLI examples for its stated workflow. No hidden execution, prompt injection, credential access, or malicious exfiltration intent was found.

1
Files scanned
261
Lines analyzed
2
Review items
0
False positives ignored
Audited by: codex

Jul 5, 2026, 03:15 PM

Most external command findings are false positives from Markdown backticks and fenced bash examples, not hidden Ruby execution. The skill intentionally depends on the belt CLI and external service links, so the remaining risk is disclosed third-party service and supply-chain exposure. No evidence of prompt injection, credential exfiltration, or obfuscated commands was found in SKILL.md.

1
Files scanned
261
Lines analyzed
4
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.

Low
Hardcoded URL
Create platform-specific video ads via [inference.sh](https://inference.sh) CLI.
The skill directs users to the external inference.sh service for video generation. This is disclosed and expected, but user prompts and media workflow data may leave the local environment.
Low
Hardcoded URL
> Requires inference.sh CLI (`belt`). [Install instructions](https://raw.githubusercontent.com/infer
The skill links to install instructions on raw.githubusercontent.com, a remote mutable source. It is not malicious by itself, but it creates a low supply-chain and dependency trust risk.
Audited by: codex

Jul 5, 2026, 03:15 PM

Most external command findings are false positives from Markdown backticks and fenced bash examples, not hidden Ruby execution. The skill intentionally depends on the belt CLI and external service links, so the remaining risk is disclosed third-party service and supply-chain exposure. No evidence of prompt injection, credential exfiltration, or obfuscated commands was found in SKILL.md.

1
Files scanned
261
Lines analyzed
4
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.

Low
Hardcoded URL
Create platform-specific video ads via [inference.sh](https://inference.sh) CLI.
The skill directs users to the external inference.sh service for video generation. This is disclosed and expected, but user prompts and media workflow data may leave the local environment.
Low
Hardcoded URL
> Requires inference.sh CLI (`belt`). [Install instructions](https://raw.githubusercontent.com/infer
The skill links to install instructions on raw.githubusercontent.com, a remote mutable source. It is not malicious by itself, but it creates a low supply-chain and dependency trust risk.
Audited by: codex

Jul 3, 2026, 01:13 PM

The skill is a Markdown guide for video ad planning and production with platform-specific specifications. Several static shell findings are real because the guide instructs users to run belt and npx commands that contact external services or install other skills; no prompt injection text was found.

1
Files scanned
261
Lines analyzed
12
Review items
0
False positives ignored
Capability review items (10)

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
> **Install the belt CLI skill:** `npx skills add belt-sh/cli`
The line instructs users to run npx skills add for another CLI skill, which is external installation activity. This is a supply-chain risk even though it is presented as setup documentation.
Medium
Ruby/shell backtick execution · 6 occurrences
```bash
The code block tells users to run belt login and belt app run commands. These commands call an external CLI service and may transmit prompts or generated media.
Medium
Ruby/shell backtick execution
```
The range includes the inline belt app store command after the related-skills block. It is a CLI command that can contact an external app store, though the risk is limited.
Low
Hardcoded URL
Create platform-specific video ads via [inference.sh](https://inference.sh) CLI.
The line links users to inference.sh and frames the skill around that external CLI service. This is a legitimate dependency, but it is still external network use.
Low
Hardcoded URL
> Requires inference.sh CLI (`belt`). [Install instructions](https://raw.githubusercontent.com/infer
The line links to raw GitHub install instructions for the required CLI. External install documentation can change and should be treated as a network and supply-chain dependency.
Audited by: codex