New: 98,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

Runtime Provisioner

by nealbridges in nealbridges/VulnHunter

VulnHunter sandbox-depth decision procedure. An agent skill from nealbridges/VulnHunter.

Apache-2.0Auto-check passedSecurity

Install Runtime Provisioner

skills CLI
$ npx skills add nealbridges/VulnHunter --skill runtime-provisioner -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install nealbridges/VulnHunter runtime-provisioner --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/nealbridges/VulnHunter.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/runtime-provisioner .claude/skills/runtime-provisioner && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
runtime-provisioner
GitHub stars
666
Token cost
~1.1k tokens
SKILL.md length
555 words
Files
1
Skills in repo
2
Repo updated
First seen
Licence
Apache-2.0

At a glance

VulnHunter sandbox-depth decision procedure. An agent skill from nealbridges/VulnHunter.

  • Works in 5 steps: Classify proof requirements → Choose the cheapest sufficient sandbox → Execute at the chosen depth → …
  • Tasks that involve Containers
  • SKILL.md covers Purpose, Placement in the methodology, Step 1 — Classify proof… and Step 2 — Choose the cheapest…, plus 5 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Runtime Provisioner is an agent skill from nealbridges/VulnHunter. VulnHunter sandbox-depth decision procedure. For each confirmed finding, classify its proof requirements, choose the cheapest sufficient sandbox (Docker container by default), record the runtime on the finding, and ensure Medium+ findings execute against the real dependency — no mocks at the boundary without an explicit, severity-capping EXECUTED-MOCK record. Use during the vulnhunt scan's Phase 2b after a finding is confirmed.

Its SKILL.md is about 1.1k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Security, covering Containers. It works with Docker. The repository describes itself as: Agentic AI security scanner that hunts exploitable vulnerabilities like an adversary, proves them with executable PoCs, and fixes them test-first. A maintained fork of Capital… The licence is Apache-2.0.

When your agent uses it

  • Tasks that involve Containers

Example prompts

  • “/runtime-provisioner”

Requirements

  • Docker

Workflow steps

5 steps, taken from the step headings in SKILL.md.

  1. Classify proof requirements
  2. Choose the cheapest sufficient sandbox
  3. Execute at the chosen depth
  4. Record the runtime on the finding
  5. Cross-check before hand-off

What it can do on your machine

Read from SKILL.md and the folder at commit 91aa73e. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Runtime Provisioner loads about 1.1k tokens when it runs. Until then it costs about 113 tokens; SKILL.md has 555 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~113
When it runs · the whole SKILL.md, loaded when a task matches
~1.1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from nealbridges/VulnHunter at commit 91aa73e, republished under its Apache-2.0 licence (© nealbridges). 555 words, ~1,100 tokens.

Download SKILL.mdSave it as .claude/skills/runtime-provisioner/SKILL.md (or your agent's skills folder).
name
runtime-provisioner
description
VulnHunter sandbox-depth decision procedure. For each confirmed finding, classify its proof requirements, choose the cheapest sufficient sandbox (Docker container by default), record the runtime on the finding, and ensure Medium+ findings execute against the real dependency — no mocks at the boundary without an explicit, severity-capping EXECUTED-MOCK record. Use during the vulnhunt scan's Phase 2b after a finding is confirmed.
trigger
/runtime-provisioner, user asks to decide sandbox depth or runtime for a finding

Runtime Provisioner — the sandbox-depth decision

Purpose

Every VulnHunter finding makes two claims: a mechanism (this input reaches this sink) and an impact (an attacker gains something they did not have). The mechanism claim is cheap to assert; the impact claim is only proven by execution against the real dependency. This procedure decides, per finding, how much runtime the impact claim needs, executes it, and records the choice so the report can show what was done to prove it — not just that something was done.

Placement in the methodology

Phase 2b, per candidate — after confirmation (Phase 2b's adversarial verify), because the sandbox decision depends on what the finding needs to prove. The decision feeds poc-impact (Phase 3): the provisioned runtime is where the PoC executes and the impact is measured.

Step 1 — Classify proof requirements

For the candidate finding, enumerate what execution the impact claim needs:

  • Boundary type — what does the sink live behind? (query grammar, HTTP client, queue, filesystem, crypto primitive, auth service, parser, ORM…)
  • Attacker-reachable surface — which attacker-controlled inputs must flow to the sink for the impact to materialize?
  • Observable outcome — what does "the attacker wins" look like? (a row returned, a file read, a request issued, a state change, a payload executed). If you cannot state the outcome in one sentence, return to the finding and restate it as an outcome before choosing a runtime.

Step 2 — Choose the cheapest sufficient sandbox

Fixed preference order; never skip a cheaper option without written cause:

  1. Docker container (default — uniform, disposable, matches CI, immune to host drift). Record the image and version tag.
  2. In-process harness, only if the dependency cannot containerize cheaply (e.g. a licensed mainframe-backed service in the target's own infra). Record the harness description and the exact dependency version pinned.
  3. Static trace only when the boundary genuinely cannot be exercised. This MUST be justified in writing on the finding — it is never the default.
Show full SKILL.md (237 more words)Show less

Step 3 — Execute at the chosen depth

Medium and above must execute in a container. No exceptions. A mock is allowed only when the real dependency is genuinely unavailable, and then the finding MUST record EXECUTED-MOCK and its severity ceiling is Low until a real-boundary run is done.

Step 4 — Record the runtime on the finding

Every finding MUST carry a Runtime line in one of these forms:

  • Runtime: docker:<image>@<tag>
  • Runtime: in-process:<description>
  • Runtime: static — <justification>
  • Runtime: EXECUTED-MOCK — <what was mocked and why>; severity ceiling Low

A finding without a Runtime line is incomplete and MUST NOT be reported.

Step 5 — Cross-check before hand-off

Before handing the finding to poc-impact, verify:

  • the Runtime line is present and well-formed;
  • the recorded runtime actually executed the boundary (not a proxy for it);
  • for Medium+: the run was in a container, or EXECUTED-MOCK is recorded with the severity ceiling noted.

The blunt sentence

A mechanism-proving assertion is not an impact measurement. "The query string contains OR WorkflowId IS NOT NULL" is the mechanism. The finding is "a victim tenant's row was returned to the attacker's listing." If your test asserts the first, it has not proven the second.

Env contract

VariableRequiredPurpose
VULNHUNT_SKILLS_DIRnoOverride the skills root this skill is loaded from.
VULNHUNT_HOST_CMDnoCommand prefix for harness session invocation, when the host harness must be driven explicitly.
VULNHUNT_DIRnoFindings directory, when operating on a prepared scan workspace.

© nealbridges, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/runtime-provisioner of nealbridges/VulnHunter.

Open the folder on GitHubat commit 91aa73e

Compare with similar skills

Runtime Provisioner next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Runtime Provisioner compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Runtime Provisioner this skillnealbridges/VulnHunter666—~1.1kAutomated safety check: PassApache-2.0
Cyberowlaikarimhabush/cyberowl263—~2.5kAutomated safety check: PassMIT
Warp Vulnerability Triagewarpdotdev/warp65k1 repos~2.1kAutomated safety check: PassAGPL-3.0
Container Scanning with GrypeAgentSecOps/SecOpsAgentKit2201 repos~2.5kAutomated safety check: PassCustom licence
Hadolint Dockerfile Security LintingAgentSecOps/SecOpsAgentKit2201 repos~4.4kAutomated safety check: PassCustom licence
Container Securityhardw00t/ai-security-arsenal104—~2.8kAutomated safety check: PassNone

Similar skills

  • Cyberowlai

    karimhabush/cyberowl

    Check if recent cybersecurity alerts from 10 international CERTs affect your current project.

    263 GitHub stars~2.5k tokensUpdated today
    SecurityAuto-check passed

  • Warp Vulnerability Triage

    warpdotdev/warp

    Gathers security findings from Dependabot, GCP container scanning, Docker Scout and Linear security issues, then triages and remediates them across Warp's repos and images.

    65k GitHub starsUsed in 1 repo~2.1k tokens
    SecurityAuto-check passed

  • Container Scanning with Grype

    AgentSecOps/SecOpsAgentKit

    Scans container images, filesystems and SBOMs with Grype for known vulnerabilities, ranks them by CVSS, EPSS and CISA KEV, and wires scans into CI/CD thresholds.

    220 GitHub starsUsed in 1 repo~2.5k tokens
    SecurityAuto-check passed

  • Hadolint Dockerfile Security Linting

    AgentSecOps/SecOpsAgentKit

    Lints Dockerfiles with Hadolint for security misconfigurations and best-practice violations, locally and in CI, with strict, balanced and permissive rule templates.

    220 GitHub starsUsed in 1 repo~4.4k tokens
    SecurityAuto-check passed

  • Container Security

    hardw00t/ai-security-arsenal

    Container and Kubernetes security assessment — image vulnerability scanning, SBOM diff analysis, K8s cluster auditing, RBAC privilege mapping, NetworkPolicy review, container escape testing, and…

    104 GitHub stars~2.8k tokensUpdated 5 mo ago
    SecurityAuto-check passed

  • Sca Trivy

    AgentSecOps/SecOpsAgentKit

    Software Composition Analysis (SCA) and container vulnerability scanning using Aqua Trivy for identifying CVE vulnerabilities in dependencies, container images, IaC misconfigurations, and license…

    220 GitHub starsUsed in 2 repos~3.7k tokens
    SecurityAuto-check passed

More from nealbridges/VulnHunter

  • Vulnhunter Run

    nealbridges/VulnHunter

    Unattended VulnHunter operator. An agent skill from nealbridges/VulnHunter.

    666 GitHub stars~711 tokensUpdated 2 days ago
    Auto-check passed

Works with

Categories

Questions about Runtime Provisioner

What does Runtime Provisioner do?

VulnHunter sandbox-depth decision procedure. An agent skill from nealbridges/VulnHunter. Runtime Provisioner is an agent skill from nealbridges/VulnHunter. VulnHunter sandbox-depth decision procedure.

When should I use Runtime Provisioner?

Runtime Provisioner fits situations like: tasks that involve Containers.

How do I install Runtime Provisioner in Claude Code?

Run `npx skills add nealbridges/VulnHunter --skill runtime-provisioner -a claude-code`. Or copy the skill folder (skills/runtime-provisioner in nealbridges/VulnHunter) into .claude/skills/runtime-provisioner in your project. Claude Code loads it when a task matches its description.

How do I install Runtime Provisioner in Codex?

Run `npx skills add nealbridges/VulnHunter --skill runtime-provisioner -a codex`. Or copy the skill folder (skills/runtime-provisioner in nealbridges/VulnHunter) into .agents/skills/runtime-provisioner in your project. Codex loads it when a task matches its description.

Can I use Runtime Provisioner in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add nealbridges/VulnHunter --skill runtime-provisioner -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/runtime-provisioner, .gemini/skills/runtime-provisioner, .github/skills/runtime-provisioner and .opencode/skills/runtime-provisioner in your project.

What does Runtime Provisioner need to run?

SKILL.md names no scripts, command-line tools or credentials: Runtime Provisioner is instructions for the agent only. Our summary lists: Docker.

Does Runtime Provisioner access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Runtime Provisioner safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Runtime Provisioner use?

Runtime Provisioner is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Runtime Provisioner use?

About 1.1k tokens (SKILL.md is roughly 4.4k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Runtime Provisioner?

Skills that share tags, products or a category with Runtime Provisioner: Cyberowlai (karimhabush/cyberowl, 263 stars), Warp Vulnerability Triage (warpdotdev/warp, 65k stars), Container Scanning with Grype (AgentSecOps/SecOpsAgentKit, 220 stars) and Hadolint Dockerfile Security Linting (AgentSecOps/SecOpsAgentKit, 220 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Runtime Provisioner?

nealbridges (a GitHub user) maintains it in nealbridges/VulnHunter, which has 666 GitHub stars. The repository holds 2 skills in this directory. The repository was last updated on October 7, 2026.

Source: nealbridges/VulnHunter on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.