Versioned security assessment

Report ID: SA-3FBCB42E

6/30/2026, 10:05:06 PM

agent-tools security assessment v2

Skill Security Certification Report

Audit History
Audit model: codex Historical report
Skill name
agent-tools
Version
v2
Maintainer
tool-belt
Source
github.com/tool-belt/skills/tree/main/tools/agent-tools/
Coverage
5 Files scanned · 596 Lines analyzed
Policy version
Unavailable

Confirmed finding summary

No confirmed security findings

The completed audit recorded no confirmed security findings. This is not proof that the Skill has no side effects.

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 flagged many shell, network, credential, and filesystem patterns. Review found most are documented infsh CLI examples for a cloud AI service, but the skill still recommends a remote shell installer and supports local file uploads to external apps. No prompt injection or confirmed malicious intent was found, so this should publish only with clear 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

5 Files scanned · 596 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.

Not recorded by this audit

Network access

May connect to external services.

Observed in 30 evidence locations

Filesystem access

May read or write local files.

Observed in 3 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 144 evidence locations

Capability review items (6)
Medium
SKILL.md:15-17references/authentication.md:5-6references/cli-reference.md:5-6
Remote Installer Piped to Shell
The documentation recommends downloading an installer and piping it directly to sh. This executes remote code in the user shell before review, even though a manual install path is also documented.
The pipe-to-shell command is directly present in multiple documentation locations. The risk is real, but the context is a normal CLI installer rather than confirmed malicious behavior.
Medium
SKILL.md:56-71references/running-apps.md:21-42
Automatic Local File Uploads to Cloud Apps
The skill documents that local file paths are automatically uploaded when used as app inputs. This is useful for media workflows but can expose local files if an agent is prompted to use sensitive paths.
The file upload behavior is explicitly described and shown with local path examples. Legitimate use is clear, but accidental disclosure remains a meaningful risk.
Medium
SKILL.md:36-53SKILL.md:49-50references/running-apps.md:130-140
Authenticated External Actions Through Cloud Apps
The skill is designed to run external cloud AI apps and includes social posting examples. These operations can create costs, upload data, or publish content outside the local workspace.
The external action capability is central to the skill and supported by concrete examples. The risk depends on user authorization and account configuration, so it is elevated but not malicious by itself.
Medium
references/authentication.md:15-16references/authentication.md:25-33references/cli-reference.md:68-72
API Key and Local Credential Handling
The documentation instructs users to log in and optionally set an API key environment variable. This is expected for the service, but prompts and logs must avoid exposing INFSH_API_KEY or stored credentials.
Credential use is clearly documented and necessary for the CLI. I did not find evidence that the skill exfiltrates secrets, but secret handling should be called out.
Low
SKILL.md:7-11SKILL.md:138-144references/app-discovery.md:108-112
Hardcoded Provider URLs Are Documentation Links
The hardcoded URLs point to inference.sh documentation, installer, distribution, and image assets. They are visible product links, not hidden callbacks or stealth endpoints.
The URLs are plainly shown in Markdown documentation. I found no evidence of covert network exfiltration endpoints.
Low
references/app-discovery.md:3-7references/cli-reference.md:19-46references/running-apps.md:3-13
Most Command Detections Are Intended CLI Examples
The many external command detections are Markdown examples or command tables for the documented infsh CLI. They are not bundled scripts that execute automatically when the skill is loaded.
The reviewed locations are documentation blocks and tables. The commands still have side effects when a user or agent runs them, so this is not a zero-risk false positive.

Risk findings

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

No confirmed security findings were recorded for this completed audit.

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.

Static false positives ignored (1)
Low
SKILL.md:20references/app-discovery.md:60references/cli-reference.md:70-72
Weak Cryptography Detections Were Not Confirmed
The reviewed locations do not implement weak cryptography. The clearest related text references SHA-256 checksum verification, which is not a weak hash for this purpose.
I inspected the flagged context and found documentation or table text, not cryptographic code. The finding appears scanner-driven rather than semantic.

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