Skills agent-tools

📦

agent-tools

Name: agent-tools
Author: toolshell

Content revision r1 Critical 🌐 Network access🔑 Env variables📁 Filesystem access⚙️ External commands

Run AI Models with inference.sh

Teams need one way to run many hosted AI apps without custom API wrappers. This skill guides inference.sh CLI discovery, authentication, execution, and result retrieval.

Supports: Claude Codex Code(CC)

⚠️ 38 Poor

Do not auto-install this skill.

The canonical policy requires operator review before any installation action.

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request

Review the Skillstore skill "agent-tools" from https://skillstore.io/skills/toolshell-agent-tools.md and its manifest at https://skillstore.io/api/skills/toolshell-agent-tools/manifest. Verify the artifact. Do not auto-install. Inspect the skill and report your findings, then wait for an operator or manual installation decision.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/toolshell-agent-tools.md Signed manifest GET /api/skills/toolshell-agent-tools/manifest Signed lockfile GET /api/skills/toolshell-agent-tools/lockfile

Test it

Using "agent-tools". Find a fast video model for a short social clip.

Expected outcome:

The assistant lists candidate video apps, chooses a fast option, explains required inputs, and asks for confirmation before execution.

Using "agent-tools". Run an image generation job and save the result.

Expected outcome:

The assistant prepares the prompt input, runs the selected image app, tracks the task, and reports the download link.

Using "agent-tools". Compare two LLM apps on a product summary task.

Expected outcome:

The assistant runs the same prompt through both apps and summarizes response quality, latency, and practical tradeoffs.

Security Audit

Critical

v5 • 7/7/2026 Open versioned report

The skill provides useful inference.sh CLI guidance, but several installer paths are confirmed risky, including curl piped to sh and remote manifest command substitution. Many markdown links and placeholder API key alerts are false positives, yet the skill still performs cloud execution and can automate X account actions. Publication should require safer installation guidance, explicit confirmation for side-effecting commands, and clearer credential handling guidance.

Files scanned

554

Lines analyzed

Review items

False positives ignored

Confirmed security concerns (6)

Critical

references/authentication.md:6

Pipe to shell pattern

curl -fsSL https://cli.inference.sh | sh

The command pipes a remote script directly into sh, which executes code before the user can inspect it. This is a critical supply-chain and remote code execution pattern.

Critical

references/authentication.md:44

Pipe to shell pattern

curl -fsSL https://cli.inference.sh | sh

The reinstall command repeats the same curl-to-sh installer pattern. Re-running a remote script through the shell preserves the critical execution risk.

Critical

references/cli-reference.md:6

Pipe to shell pattern

curl -fsSL https://cli.inference.sh | sh

The CLI reference installs by piping a remote URL directly to sh. This executes unreviewed remote code and is a critical installer anti-pattern.

Critical

SKILL.md:16

Pipe to shell pattern

curl -fsSL https://cli.inference.sh | sh

The main install instructions pipe a remote URL directly into sh. This executes unreviewed network content and should be treated as a critical blocker.

High

SKILL.md:49-50 SKILL.md:78

External Account Actions Without Guardrails

The skill advertises X automation and includes a direct post example. A connected account could publish or interact publicly unless callers require explicit confirmation.

The cited lines show both a concrete post command and a broader X automation category. This is clear external account side-effect capability, not just a documentation link.

Medium

SKILL.md:9 references/running-apps.md:119-121

Cloud Submission of User Inputs

The core workflow runs prompts and files through inference.sh cloud apps, and file outputs are returned as cloud URLs. Sensitive inputs may leave the local workspace.

The skill explicitly describes cloud execution and the running-apps reference states file outputs are delivered through URLs. This supports a data exposure concern when users submit sensitive content.

Capability review items (18)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

High

references/cli-reference.md:84

Hidden file in home directory

infsh completion fish > ~/.config/fish/completions/infsh.fish

The command writes generated shell completion content under ~/.config in the user home directory. This is expected CLI setup, but it still modifies hidden shell configuration state.

High

SKILL.md:23-32

Ruby/shell backtick execution

> ```bash

The manual install block downloads remote artifacts, parses a manifest, extracts an archive, and moves a binary into PATH. This is real external command and supply-chain risk.

High

SKILL.md:26

Shell command substitution

> curl -LO $(curl -fsSL https://dist.inference.sh/cli/manifest.json | grep -o '"url":"[^"]*"' | grep

Line 26 uses command substitution fed by remote manifest data to construct a download URL. This creates unpinned remote artifact selection and shell parsing risk.

High

SKILL.md:23-32

Template literal with command substitution

> ```bash

The referenced block includes command substitution and remote binary installation steps. It is documentation, but the shell workflow is risky if followed automatically.

High

SKILL.md:31

Hidden file in home directory

> mv inferencesh-cli-* ~/.local/bin/inferencesh

The manual install moves a downloaded binary into ~/.local/bin. That modifies a hidden home path on PATH and can persist executable code.

Medium

references/cli-reference.md:84

Hidden file access

infsh completion fish > ~/.config/fish/completions/infsh.fish

The fish completion command writes to a hidden home configuration path. The behavior is documented, but it is a real filesystem side effect that should be explicit.

Medium

SKILL.md:15-18

Ruby/shell backtick execution

```bash

The fenced shell block contains setup commands including remote installation and login. The backtick token is markdown, but the surrounding block presents executable shell steps.

Medium

SKILL.md:36-54

Ruby/shell backtick execution

```bash

The examples block contains infsh app run commands that submit prompts to cloud AI apps and includes an X post example. These commands can have account and network side effects.

Medium

SKILL.md:65-66

Ruby/shell backtick execution

| Run app | `infsh app run google/veo-3-1-fast --input input.json` |

This command runs a named cloud app using an input file, which may transmit data and consume account resources. It is intended behavior but has real side-effect risk.

Medium

SKILL.md:66-67

Ruby/shell backtick execution

| Run without waiting | `infsh app run <app> --input input.json --no-wait` |

The no-wait command starts a cloud task without waiting for completion. Accidental execution can create external work or spend account credits.

Medium

SKILL.md:83-109

Ruby/shell backtick execution

```bash

The block recommends npx skills add commands that install additional skills from a remote repository. Installing extra skills changes the agent environment and expands trusted instructions.

Medium

SKILL.md:31

Hidden file access

> mv inferencesh-cli-* ~/.local/bin/inferencesh

The command writes an installed binary under a hidden home directory path. This is an intended install step, but it is still a real filesystem modification.

Low

references/authentication.md:6

Hardcoded URL

curl -fsSL https://cli.inference.sh | sh

Line 6 fetches a remote installer from cli.inference.sh, so it is a real external network dependency. The stronger execution risk is covered by the pipe-to-shell finding.

Low

references/authentication.md:44

Hardcoded URL

curl -fsSL https://cli.inference.sh | sh

Line 44 repeats the remote installer URL for reinstalling the CLI. This is a real network dependency even though the execution risk is adjudicated separately.

Low

references/cli-reference.md:6

Hardcoded URL

curl -fsSL https://cli.inference.sh | sh

The installation command retrieves a remote script from cli.inference.sh. This is a real network dependency even though the pipe-to-shell finding captures the higher risk.

Low

SKILL.md:16

Hardcoded URL

curl -fsSL https://cli.inference.sh | sh

The installation example fetches a remote script from cli.inference.sh. This is a real network dependency, with execution risk covered by the pipe-to-shell finding.

Low

SKILL.md:25

Hardcoded URL

> curl -LO https://dist.inference.sh/cli/checksums.txt

The manual installation instructions download a checksums file from a remote distribution host. This is a real network dependency during setup.

Low

SKILL.md:26

Hardcoded URL

> curl -LO $(curl -fsSL https://dist.inference.sh/cli/manifest.json | grep -o '"url":"[^"]*"' | grep

The command downloads a URL derived from a remote manifest. This is a real network dependency with supply-chain relevance.

Risk Factors

🌐 Network access (17)

references/authentication.md:6 references/authentication.md:44 references/cli-reference.md:6 references/running-apps.md:129 SKILL.md:7 SKILL.md:11 SKILL.md:16 SKILL.md:20 SKILL.md:25 SKILL.md:26 SKILL.md:120 SKILL.md:121 SKILL.md:122 SKILL.md:123 SKILL.md:124 SKILL.md:125 SKILL.md:126

🔑 Env variables (3)

references/authentication.md:30 references/authentication.md:53 references/cli-reference.md:72

📁 Filesystem access (4)

references/cli-reference.md:84 references/cli-reference.md:84 SKILL.md:31 SKILL.md:31

⚙️ External commands (18)

SKILL.md:15-18 SKILL.md:18-20 SKILL.md:20-23 SKILL.md:23-32 SKILL.md:32-36 SKILL.md:36-54 SKILL.md:54-60 SKILL.md:60-61 SKILL.md:61-62 SKILL.md:62-63 SKILL.md:63-64 SKILL.md:64-65 SKILL.md:65-66 SKILL.md:66-67 SKILL.md:67-83 SKILL.md:83-109 SKILL.md:26 SKILL.md:23-32

Detected Patterns

Pipe to shell pattern×4

Audited by: codex View Audit History →

Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report

Copy report link

https://skillstore.io/skills/toolshell-agent-tools/audits/5?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/toolshell-agent-tools/security.svg)](https://skillstore.io/skills/toolshell-agent-tools?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/toolshell-agent-tools?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/toolshell-agent-tools/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/toolshell-agent-tools.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

Academic citations (APA · BibTeX · CFF)

APA citation

toolshell. (2026). agent-tools security audit report (audit version 5) [Author version unspecified]. Skillstore. https://skillstore.io/skills/toolshell-agent-tools/audits/5

BibTeX citation

@techreport{toolshell-toolshell-agent-tools-2026,
  author = {toolshell},
  title = {agent-tools security audit report (audit version 5)},
  institution = {Skillstore},
  year = {2026},
  number = {5},
  url = {https://skillstore.io/skills/toolshell-agent-tools/audits/5},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "agent-tools security audit report (audit version 5)"
version: "unspecified"
type: report
authors:
  - name: "toolshell"
date-released: "2026-07-07"
url: "https://skillstore.io/skills/toolshell-agent-tools/audits/5"
identifiers:
  - type: other
    value: "skillstore:toolshell-agent-tools:audit:5"
    description: "Skillstore immutable audit report identifier"

Compare variants

20 installable variants

Each author remains a separate installable skill. The recommended variant is ranked by Skillstore evidence.

Why this variant is first

Higher Skillstore usage

Author / source Skillstore Score Evidence Confidence Skillstore usage Updated

tul-sh Recommended

tul-sh-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 28

Updated

2026-08-21

toolshell Current

toolshell-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 25

Updated

2026-08-21

inferen-sh

inferen-sh-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 17

Updated

2026-08-21

tool-belt

tool-belt-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 14

Updated

2026-08-21

inferencesh

inferencesh-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 13

Updated

2026-08-21

inference-sh-skills

inference-sh-skills-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 12

Updated

2026-08-21

inference-sh-7

inference-sh-7-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 12

Updated

2026-08-21

inf-sh

inf-sh-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 11

Updated

2026-08-21

inference-skills

inference-skills-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 11

Updated

2026-08-21

inference-shell

inference-shell-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 11

Updated

2026-08-21

inference-sh-9

inference-sh-9-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 11

Updated

2026-08-21

qu-skills

qu-skills-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 10

Updated

2026-08-21

inference-sh-6

inference-sh-6-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 10

Updated

2026-08-21

inference-sh-3

inference-sh-3-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 10

Updated

2026-08-21

infsh-skills

infsh-skills-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 9

Updated

2026-08-21

inference-sh-0

inference-sh-0-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 9

Updated

2026-08-21

skillssh

skillssh-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 8

Updated

2026-08-21

inference-sh-8

inference-sh-8-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 7

Updated

2026-08-21

101-skills

101-skills-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 7

Updated

2026-08-21

halt-catch-fire

halt-catch-fire-agent-tools

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 6

Updated

2026-08-21

Skillstore Score

Why this score Evidence Confidence: Medium

Architecture

Maintainability

Content

Community

Spec Compliance

What You Can Build

Prototype AI Media Workflows

Find image, video, audio, or 3D apps and run quick experiments from the CLI.

Compare Hosted Models

Run similar prompts across LLM or media apps and review the outputs in one workflow.

Operate Content Tasks

Submit cloud generation jobs, track task status, and collect downloadable output links.

Try These Prompts

Find an App

Help me find inference.sh apps for [task]. Show likely categories, app names, and what input each app requires.

Run One Job

Use inference.sh to run [app name] for [goal]. Prepare the input, run the job, and summarize the result.

Build a Workflow

Plan and run an inference.sh workflow that creates [asset], improves it with [tool], and saves the final result links.

Evaluate Several Apps

Compare [models or apps] on [criteria]. Generate sample inputs, run each app, collect outputs, and summarize tradeoffs.

Best Practices

Inspect app schemas before sending production data.
Use input files for complex runs, and review them before execution.
Confirm any social posting, paid run, or public output before running the CLI.

Avoid

Do not paste private keys, credentials, or unreleased content into model inputs.
Do not pipe remote installers into a shell in unattended environments.
Do not use X automation without review, rate limits, and account owner approval.