security-audit-platform

A repository-wide security audit process for MCP Runtime, covering its components and trust boundaries. MCP Runtime is the system being reviewed, and a trust boundary is a place where data or permissions cross between parts of it.

In plain words
What is it for?
Building a threat model, reviewing access controls and tenant separation, testing protocol handling, checking audit logs and TLS, and running security tests against a live cluster.
Why use it?
It gives a structured way to look for security problems across the whole platform rather than in one change.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/agent-hellboy/mcp-runtime/security-audit-platform
Any agent
npx skills add Agent-Hellboy/mcp-runtime --skill security-audit-platform
Clone the repo
git clone --depth 1 https://github.com/Agent-Hellboy/mcp-runtime

Made for: Claude Code, Codex.

Per session 123 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,289 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 3 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00123 $0.03289
Opus 5 $0.00062 $0.01644
Sonnet 5 $0.00025 $0.00658
Haiku 4.5 $0.00012 $0.00329

Measured 2d ago against content hash 22d9746d1e54, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

security-audit-platform scanned grade B with 3 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Sends data to an external URLlowData exfiltration

A POST to an outside endpoint may be telemetry or may be exfiltration; either way the mod talks to somewhere, and you should know where.

- **Large body**: `dd if=/dev/urandom bs=1M count=200 | curl -X POST --data-binary @- ...`;

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Cloud metadata endpointmediumServer-side request forgery

One request to 169.254.169.254 can return temporary IAM credentials.

`127.0.0.1`, `169.254.169.254`, `[::1]`, IPv6 zone IDs, schemes other than

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

got="$(curl -sS -o /dev/null -w "%{http_code}\n" -X "$method" "${headers[@]}" "$BASE$path")"
.codex/skills/security-audit-platform/SKILL.md · 268 lines

How it starts

The opening of the file, as written. The whole thing — 268 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Security Audit — Platform-wide

Overview

Use this skill when the goal is a deep, repository-wide security assessment of MCP Runtime, not a PR review. The output is a structured report with a threat model, a per-endpoint authn/authz matrix, tenant-isolation probes, protocol fuzz results, audit-log integrity tests, TLS hygiene, and DAST against a live cluster. Findings use the shared template at ../_shared/FINDINGS-TEMPLATE.md.

This skill is intentionally heavy. Expect hours, not minutes. Skip nothing silently — every check that did not run becomes a recorded gap.

Step 1 — Build the threat model before running tools

Produce a STRIDE table per component. Components to cover:

  • operator (cmd/operator/, internal/operator/): reconciles MCPServer, MCPAccessGrant, MCPAgentSession; injects gateway sidecar.
  • mcp-gateway (services/mcp-gateway/): in-pod sidecar enforcing rendered policy, emitting audit events.
  • platform-api (services/platform-api/): identity, admin, registry forward-auth.
  • runtime-api (services/runtime-api/): runtime governance, deployments, registry push.
  • analytics-api (services/analytics-api/): ClickHouse events and usage analytics.
  • sentinel-ui (services/ui/): browser UI, login, dashboards.
  • sentinel-ingest (services/ingest/): high-volume event intake.
  • sentinel-processor (services/processor/): event processing, ClickHouse writes.
  • registry (k8s/, config/): Distribution v2 registry (HTTP dev or HTTPS prod).
  • traefik plugins (traefik-plugins/): PII redactor, dev-only middleware.
  • CRD types (api/v1alpha1/): trust source for resource shapes.
  • CI (.github/workflows/): build, sign, test, release pathways.

For each component fill: Spoofing, Tampering, Repudiation, Information disclosure, Denial of service, Elevation of privilege. State preconditions, attacker, and impact.

Attacker profiles to enumerate:

  • anon-public: someone reaching platform.<domain> from the internet.
  • anon-cluster: a pod inside the cluster with no creds.
  • authenticated-user: holds UI_API_KEY or a logged-in session.
  • admin-user: holds an ADMIN_API_KEYS value.
  • ingest-only: holds an INGEST_API_KEYS value.
  • rogue-mcp-image: a malicious image pulled from the registry.
  • rogue-tenant-agent: a session for tenant A trying to reach tenant B.
  • mitm-ingress / mitm-registry: someone between client and ingress.
  • compromised-CI-token: a leaked GH Actions secret.

Read the full file on GitHub · 268 lines

Files

What ships with it

3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 268 lines · 123 tokens per session scan B 22d9746d1e54

Subscribe to this mod's changes

security-audit-platform is a skill published in the GitHub repository Agent-Hellboy/mcp-runtime (5 stars, last pushed 8d ago), licensed Apache-2.0. It adds 123 tokens to every session and 3,289 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it B with 3 findings (sends data to an external url, cloud metadata endpoint, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

local-frontend-check

Smoke-test or verify UI behaviour on the local Jarvis Registry frontend running at http://localhost/gateway. Use for manual regression checks, bug-fix verification, and end-to-end confirmation of specific flows without running the automated test suite.

ascending-llc/jarvis-registry · 52 tokens

release-notes

Create release notes for a new version tag. Gathers all commits, PRs, issues fixed, and breaking changes since a previous release. Creates the release notes markdown file, tags the repo, and pushes. Asks the user to confirm the base version to diff against.

agentic-community/mcp-gateway-registry · 59 tokens

agentcore-register

Given an MCP server URL, probe the server via curl to discover its metadata and tools, then generate a markdown file with copy-pasteable content for each field in the Amazon Bedrock AgentCore "Create record" form.

agentic-community/mcp-gateway-registry · 49 tokens

debug

Debug issues in the MCP Gateway Registry using first-principles thinking. Invoke when something is broken, timing out, returning errors, or behaving unexpectedly. Forces structured root-cause analysis before any code change is proposed.

agentic-community/mcp-gateway-registry · 45 tokens

search-benchmark

Generate a search quality benchmark for the AI Registry. Generates ground truth from the registry's assets, runs 100+ queries against the semantic search API, evaluates results using NDCG@10/MRR/Recall, and produces a markdown report. Use when you want to measure search quality after changes to the scoring algorithm…

agentic-community/mcp-gateway-registry · 75 tokens

benchmark-report

Generate a benchmark report from stress test results (registration, API performance, search concurrency). Reads JSON result files and produces a markdown report suitable for docs/benchmarks/.

agentic-community/mcp-gateway-registry · 36 tokens