This report does not block or authorize the manifest or ZIP.
Static analysis reported many high-risk patterns, but review found no evidence of malicious intent, prompt injection, credential exfiltration, weak cryptography, or filesystem traversal. Most flagged items are Markdown examples, Azure CLI guidance, a relative documentation link, or helper scripts that generate an Azure AI Foundry portal URL. The remaining risk is medium because the skill guides authenticated Azure deployment actions that can create billable cloud resources.
Report position
Historical report
Open audit history before using this report to install.
Audit attestation
Not attestable
The required immutable binding is incomplete.
Human verification
Not verified
No human verification is recorded for this report.
Coverage
4 Files scanned · 388 Lines analyzed
4 items shown for review
Limitations
This report does not claim runtime or sandbox execution and does not prove the absence of side effects.
01
Evidence chain
Follow the evidence from source binding to the install contract. Available evidence supports verification; it is not a safety guarantee.
The skill routes users through Azure OpenAI deployment workflows and requires confirmation before deployment. This is legitimate for the skill purpose, but it can create billable Azure resources when used with authenticated credentials.
The skill explicitly covers deployment mode selection, project resolution, and pre-deployment validation. The confirmation requirement reduces risk, but authenticated cloud changes remain operationally sensitive.
The Bash helper accepts subscription, resource group, project, and deployment names, then uses shell command substitution to encode the subscription ID and print a portal URL. The arguments are quoted and no command injection was confirmed, but identifier validation would reduce malformed URL risk.
The static command-substitution finding is real, but the semantic use is limited to local string transformation and URL output. No eval, remote request, or unquoted command construction was found.
03
Risk findings
Confirmed security concerns are separated from items that still need review.
Markdown Relative Link Is Not Filesystem Traversal
The path traversal alerts correspond to a Markdown link that points readers to a quota skill document. No evidence found of runtime file access using that relative path.
The flagged sequence appears inside documentation text as a relative link. It is not used by code that reads, writes, or executes files.
The system reconnaissance alert points to a negative test case for an invalid Azure region. No evidence found of host, network, or environment reconnaissance instructions.
The cited line is part of a table of expected failure cases. It tests graceful handling of an invalid region rather than collecting system information.
04
Expert evidence
Immutable subject identity, scanner metadata, dismissed matches, and source-level evidence.
Artifact subject
Marketplace commit
Unavailable
Content hash
Unavailable
Tree hash
Unavailable
Skill path
Unavailable
Audit payload hash
Unavailable
Analysis metadata
Audit model: codex
Analysis state: Complete
Scope is limited to the recorded files, lines, methods, and evidence. No runtime or sandbox execution is claimed.
The hardcoded URL points to the Azure AI Foundry portal and is only concatenated into a link that is printed or returned. No evidence found of data being sent to that URL by the scripts.
Both scripts build a portal URL as output rather than performing a network request. The destination is an expected Microsoft Azure domain for this deployment skill.
The weak cryptography alerts occur in descriptive deployment text, headings, and prerequisite sections. No evidence found of MD5, SHA1, DES, RC4, or other weak cryptographic algorithms being used.
Manual review of the cited lines found Azure OpenAI deployment instructions, not cryptographic API use. The scanner likely matched substrings in ordinary documentation.