📦

Audit History

content-gap-analysis - 4 audits

Version comparison

Capability and finding changes across audited versions, newest first.

VersionDateResultReview itemsChange vs previous
v4 LatestJul 12, 2026, 11:31 AM No confirmed findings1No capability change
v3 Jul 12, 2026, 11:31 AM No confirmed findings1No capability change
v2 Jul 6, 2026, 04:06 PM 1 confirmed2No capability change
v1 Jul 4, 2026, 03:53 PM No confirmed findings0Baseline

Jul 12, 2026, 11:31 AM

Most static alerts are false positives caused by Markdown fences, metadata URLs, relative documentation links, and ordinary scoring language. The command at SKILL.md:50 is a genuine medium-risk external command because a user-derived domain placeholder is passed without explicit validation or shell-safe argument handling.

4
Files scanned
384
Lines analyzed
4
Review items
0
False positives ignored
Capability review items (1)

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
**Keyless competitor-coverage inventory**: `python3 "${CLAUDE_PLUGIN_ROOT}/scripts/connectors/firecr
The line explicitly instructs execution of a Python connector and leaves the user-derived competitor domain unquoted. Without strict validation and argument-safe invocation, crafted input could alter command arguments.
Audited by: codex

Jul 12, 2026, 11:31 AM

Most static alerts are false positives caused by Markdown fences, metadata URLs, relative documentation links, and ordinary scoring language. The command at SKILL.md:50 is a genuine medium-risk external command because a user-derived domain placeholder is passed without explicit validation or shell-safe argument handling.

4
Files scanned
384
Lines analyzed
4
Review items
0
False positives ignored
Capability review items (1)

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
**Keyless competitor-coverage inventory**: `python3 "${CLAUDE_PLUGIN_ROOT}/scripts/connectors/firecr
The line explicitly instructs execution of a Python connector and leaves the user-derived competitor domain unquoted. Without strict validation and argument-safe invocation, crafted input could alter command arguments.
Audited by: codex

Jul 6, 2026, 04:06 PM

Most static findings are false positives from Markdown code fences, fixed memory paths, metadata URLs, and documentation links. The confirmed risk is the optional connector workflow, which asks agents to run local scripts and perform outbound competitor scraping.

4
Files scanned
384
Lines analyzed
6
Review items
0
False positives ignored

Confirmed security concerns (1)

Medium
Optional Connector Workflow Invokes Outbound Scraping
The skill directs agents to use trend and Firecrawl connector scripts for competitor coverage and page scraping. This can trigger outbound requests and send user-provided domains or topics outside the chat if run without confirmation.
Lines 48-50 explicitly describe connector-based trend scouting and Firecrawl scraping. The workflow appears legitimate, but the network and third-party data-transfer behavior is clear.
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
**Trend-scout as a gap-discovery input (keyless)**: feed the multi-source trend scout — Google Trend
The text directs agents to use a connector script for trend scouting. It is a legitimate workflow, but it expands the skill into local script execution and outbound data collection.
Medium
Ruby/shell backtick execution
**Keyless competitor-coverage inventory**: `python3 "${CLAUDE_PLUGIN_ROOT}/scripts/connectors/firecr
Line 50 includes a python3 command that runs firecrawl.py with user-supplied domain and topic values. This is real external command execution with network scraping behavior.
Audited by: codex

Jul 4, 2026, 03:53 PM

All static findings are false positives after contextual review. The flagged items are Markdown prompt fences, repository documentation links, homepage metadata, bounded workspace paths, or optional connector instructions; no prompt injection, credential exfiltration, arbitrary command execution, or malicious traversal intent was found.

4
Files scanned
384
Lines analyzed
3
Review items
0
False positives ignored
Audited by: codex