Versioned security assessment

Report ID: SA-5E797438

6/30/2026, 8:04:42 AM

microsoft-foundry security assessment v2

Skill Security Certification Report

Audit History
Audit model: codex Historical report
Skill name
microsoft-foundry
Version
v2
Maintainer
microsoft
Source
github.com/microsoft/github-copilot-for-azure/tree/main/plugin/skills/microsoft-foundry/
Coverage
22 Files scanned · 3,158 Lines analyzed
Policy version
Unavailable

Highest confirmed finding severity

Low

1 confirmed security finding requires attention.

Installation context

Historical evidence

This report may not describe the currently installable artifact. Open the current Skill page for install guidance.

Open current Skill page

This report does not block or authorize the manifest or ZIP.

Static analysis produced a very high score because the skill contains many Markdown command examples, Azure URLs, relative links, and environment variable references. AI review found no prompt injection attempt or clear malicious intent, but the skill intentionally guides cloud provisioning, Docker builds, role assignments, key retrieval, and environment variable handling, so it should publish with operational risk warnings.

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

22 Files scanned · 3,158 Lines analyzed

6 items shown for review

Limitations

This report does not claim runtime or sandbox execution and does not prove the absence of side effects.

Evidence chain

Follow the evidence from source binding to the install contract. Available evidence supports verification; it is not a safety guarantee.

  1. Source

    Binding unavailable

  2. Artifact

    Identity incomplete

  3. Audit

    Complete

  4. Install contract

    Open manifest to verify

    Open manifest

Capabilities observed

Observed means this report recorded supporting evidence. Not recorded does not prove that a capability is absent.

Contains scripts

May execute code included with the Skill.

Observed in 3 evidence locations

Network access

May connect to external services.

Observed in 3 evidence locations

Filesystem access

May read or write local files.

Observed in 2 evidence locations

Env variables

May read values from the process environment.

Observed in 3 evidence locations

External commands

May invoke commands or programs outside the Skill.

Observed in 5 evidence locations

Capability review items (5)
Medium
project/create/create-foundry-project.md:18-57resource/create/create-foundry-resource.md:21-82foundry-agent/deploy/deploy.md:10-88quota/quota.md:24-65
Cloud Infrastructure Commands Require User Review
The skill intentionally guides Azure CLI, Azure Developer CLI, Docker, and ACR operations that can create, modify, deploy, or delete cloud resources. This is legitimate for a Microsoft Foundry operations skill, but the commands can affect billing, access, and production workloads if run without confirmation.
The commands are documented workflow steps with clear Azure operational intent, not hidden execution. Risk remains meaningful because they can provision, deploy, delete, or change paid cloud resources.
Medium
foundry-agent/deploy/deploy.md:59-88foundry-agent/create/agent-framework/SKILL.md:91-97references/sdk/foundry-sdk-py.md:49-66
Environment Variables And Secrets Are Handled During Deployment
The deployment workflow scans source code for environment variables, pre-fills values from azd configuration, creates local environment files, and includes environment variables in hosted agent payloads. Sensitive values are expected to be masked, but mishandling could expose credentials or long-lived configuration secrets.
The context confirms legitimate deployment configuration handling and explicit masking guidance. The risk is not malicious, but credentials and environment values are part of the workflow.
Medium
rbac/rbac.md:35-121
RBAC And Service Principal Workflows Can Create Privileged Credentials
The RBAC workflow includes role assignment, managed identity configuration, and service principal creation for CI/CD. These actions are legitimate, but generated credentials and broad roles require least-privilege review and secure storage outside chat history.
The workflow clearly describes identity and permission management, including service principal setup. The concern is operational exposure from credential creation and role assignment rather than malicious behavior.
Low
SKILL.md:95-97project/create/create-foundry-project.md:123quota/quota.md:27-47
Official Documentation And Azure API URLs Are Hardcoded
The network findings are primarily links to Microsoft documentation, Azure portal pages, GitHub samples, and Azure management endpoints used for quota checks. These URLs support the skill purpose and do not indicate data exfiltration.
The reviewed URLs are expected Azure and Microsoft references or management APIs. I did not find evidence of suspicious third-party collection endpoints.
Low
foundry-agent/deploy/deploy.md:174-206resource/create/create-foundry-resource.md:61
Relative Links Triggered Path Traversal Heuristics
The filesystem path traversal findings appear to come from relative Markdown links between skill documents. These links help route readers to related workflow files and do not access arbitrary filesystem paths.
The evidence points to documentation-relative navigation rather than filesystem reads or writes. The scanner found traversal tokens, but not an exploit path.

Risk findings

Confirmed security concerns are separated from items that still need review.

Confirmed security concerns (1)

RISK-001 Low
SKILL.md:27-33foundry-agent/create/agent-framework/references/workflow-basics.md:11-26foundry-agent/create/agent-framework/references/agent-samples.md:8-29
Static Analyzer Matched Documentation Syntax As Dangerous Code
Many Ruby backtick, weak cryptography, and command substitution findings are false positives caused by Markdown inline code, code fences, headings, or normal shell examples. No evidence found that the skill hides executable code or performs automatic command execution by itself.
The sampled locations are Markdown instructions and examples, not source code loaded by the skill runtime. The pattern labels do not match the semantic context.

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.

Verify and export

JSON report CITATION.cff SARIF 2.1.0 Signed installation manifest

The manifest and lockfile bind install artifacts to cryptographic hashes. This integrity claim is separate from the security assessment.

Audit attestation: not_attestable