📦

Audit History

edge-computing-patterns - 10 audits

Version comparison

Capability and finding changes across audited versions, newest first.

VersionDateResultReview itemsChange vs previous
v10 LatestJul 23, 2026, 06:13 AM 2 confirmed0No capability change
v9 Jul 7, 2026, 05:39 PM No confirmed findings0No capability change
v8 Jul 6, 2026, 04:18 AM No confirmed findings0External commands
v7 Jun 28, 2026, 09:57 AM 2 confirmed1No capability change
v6 Jan 21, 2026, 04:16 PM No confirmed findings0 External commands
v5 Jan 16, 2026, 04:40 PM No confirmed findings0No capability change
v4 Jan 16, 2026, 04:40 PM No confirmed findings0External commandsNetwork accessEnv variables
v3 Jan 10, 2026, 10:35 AM No confirmed findings0No capability change
v2 Jan 10, 2026, 10:35 AM No confirmed findings0No capability change
v1 Jan 10, 2026, 10:35 AM No confirmed findings0Baseline

Jul 23, 2026, 06:13 AM

All 34 static findings are false positives caused by documentation, expected edge APIs, TypeScript syntax, and public runtime metadata. No command execution, secret-file access, reconnaissance, prompt injection, or malicious network intent appears. Semantic review found shared-cache exposure risk and a non-atomic KV rate limiter in instructional examples.

1
Files scanned
234
Lines analyzed
5
Review items
0
False positives ignored

Confirmed security concerns (2)

High
Unsafe Shared-Cache Examples
The examples cache origin responses by request URL without excluding authenticated, private, or unsuccessful responses. This can expose personalized content across users.
Both examples write fetched responses into a shared cache without checking authorization, cookies, private directives, or public cache eligibility.
Medium
Non-Atomic Distributed Rate Limiting
The KV read-modify-write counter is non-atomic and eventually consistent. Concurrent or distributed requests can undercount and bypass the intended limit.
The example performs separate KV get and put operations with no atomic increment or serialization mechanism.
Audited by: codex

Jul 7, 2026, 05:39 PM

All static findings are false positives caused by Markdown code fences, inline API names, and illustrative TypeScript examples. No prompt injection, executable scripts, command execution, data exfiltration, or malicious network behavior was found in SKILL.md.

1
Files scanned
234
Lines analyzed
3
Review items
0
False positives ignored
Audited by: codex

Jul 6, 2026, 04:18 AM

The static findings are false positives caused by Markdown code fences, inline code formatting, and edge runtime examples. No prompt injection, command execution intent, credential exfiltration, or malicious network destination was found in SKILL.md.

1
Files scanned
234
Lines analyzed
3
Review items
0
False positives ignored
Audited by: codex

Jun 28, 2026, 09:57 AM

Static external command and critical combination alerts are false positives from Markdown code fences, template literals, and a sentence that lists child_process as unavailable. The skill is documentation-only, with legitimate edge runtime examples that use fetch and environment bindings. No prompt injection attempt, credential exfiltration, executable installer, or malicious intent was found.

1
Files scanned
234
Lines analyzed
5
Review items
1
False positives ignored

Confirmed security concerns (2)

Low
Network Calls Are Expected Edge Runtime Examples
The fetch calls demonstrate proxying, authentication forwarding, rate limiting, and cache fill behavior. They do not target suspicious endpoints or send credentials to unknown domains.
The network API usage is real, but it appears in instructional edge application examples. No evidence shows exfiltration, hidden endpoints, or malicious network behavior.
Low
Environment Variable Usage Is Not Exfiltration
The examples read VERCEL_REGION for metadata and JWT_SECRET for token verification. The values are not logged, transmitted to a third party, or written to storage.
The environment access is visible and limited to normal edge runtime configuration. It includes a secret example, but the secret is only used for local JWT verification.
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.

Low
Cache Examples Need Adaptation For Authenticated Routes
The cache examples store origin responses after fetch calls. Users should avoid copying this pattern into personalized or authenticated routes without cache key controls.
The examples are generic documentation, so impact depends on how users adapt them. The caching pattern can become unsafe if used for personalized responses.
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.

Low
Static External Command Alerts Are Documentation False Positives
The reported command execution matches are Markdown code blocks, template literals, and a note that Node.js child_process is unavailable. No executable command invocation was found.
The referenced lines are code fence delimiters, TypeScript examples, or text describing unavailable APIs. There is no script, shell command, or runtime hook in the skill.
Audited by: codex

Jan 21, 2026, 04:16 PM

Static analysis flagged patterns in documentation code examples. All findings are false positives. The skill contains educational content about edge computing deployment with example code snippets in markdown. No actual command execution, credential exfiltration, or malicious behavior exists. Network access and env references are documentation references only.

2
Files scanned
901
Lines analyzed
2
Review items
0
False positives ignored

Risk Factors

Audited by: claude

Jan 16, 2026, 04:40 PM

AI analysis failed after multiple attempts - MANUAL REVIEW REQUIRED before publishing. This skill cannot be auto-published until reviewed by a human.

2
Files scanned
416
Lines analyzed
3
Review items
0
False positives ignored

Detected Patterns

Node.js child_process moduleHardcoded URLWeak cryptographic algorithmSystem reconnaissanceRuby/shell backtick executionFetch API callEnvironment variable access (dot notation)Environment variable objectEnvironment file access[HEURISTIC] DANGEROUS COMBINATION: Code execution + Network + Credential access
Audited by: claude

Jan 16, 2026, 04:40 PM

AI analysis failed after multiple attempts - MANUAL REVIEW REQUIRED before publishing. This skill cannot be auto-published until reviewed by a human.

2
Files scanned
416
Lines analyzed
3
Review items
0
False positives ignored

Detected Patterns

Node.js child_process moduleHardcoded URLWeak cryptographic algorithmSystem reconnaissanceRuby/shell backtick executionFetch API callEnvironment variable access (dot notation)Environment variable objectEnvironment file access[HEURISTIC] DANGEROUS COMBINATION: Code execution + Network + Credential access
Audited by: claude

Jan 10, 2026, 10:35 AM

This is a documentation-only skill containing educational content about edge computing patterns. No executable code, scripts, network calls, or file system access. All content is legitimate educational material.

1
Files scanned
234
Lines analyzed
0
Review items
0
False positives ignored
No confirmed security findings were recorded for this completed audit.
Audited by: claude

Jan 10, 2026, 10:35 AM

This is a documentation-only skill containing educational content about edge computing patterns. No executable code, scripts, network calls, or file system access. All content is legitimate educational material.

1
Files scanned
234
Lines analyzed
0
Review items
0
False positives ignored
No confirmed security findings were recorded for this completed audit.
Audited by: claude

Jan 10, 2026, 10:35 AM

This is a documentation-only skill containing educational content about edge computing patterns. No executable code, scripts, network calls, or file system access. All content is legitimate educational material.

1
Files scanned
234
Lines analyzed
0
Review items
0
False positives ignored
No confirmed security findings were recorded for this completed audit.
Audited by: claude