knowledge-mcp: Agent for Claude Code

.claude/agents/infra-reviewer.md

infra-reviewer is an agent for Claude Code from fulminate-io/knowledge-mcp. It costs 76 tokens per session (4,748 once invoked), scanned A, original, Apache-2.0.

An infrastructure-change reviewer that checks configuration and planned changes before deployment, provisioning, upgrades, or image releases. It compares claims against available evidence such as provider documentation, source code, and live cloud state.

In plain words
What is it for?
Use it to audit infrastructure changes before running commands that alter cloud resources or deploy new versions.
Why use it?
It helps expose configuration that looks correct but could fail when applied to real infrastructure. Unverified values are marked as findings instead of being treated as facts.

Agent for Claude Code

Written for Claude Code: installed under .claude/. Also seen: model in frontmatter; reads .claude/ paths.

This is fulminate-io/knowledge-mcp's own configuration. It tells Claude Code how to work on knowledge-mcp itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything knowledge-mcp configures →

Reuse

Borrowing it

Nothing to install: this file belongs to fulminate-io/knowledge-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/fulminate-io/knowledge-mcp/main/.claude/agents/infra-reviewer.md
Clone the repo
git clone --depth 1 https://github.com/fulminate-io/knowledge-mcp

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for infra-reviewer

README.md
[![agentmods](https://agentmods.dev/badge/agents/fulminate-io/knowledge-mcp/infra-reviewer/github.svg)](https://agentmods.dev/agents/fulminate-io/knowledge-mcp/infra-reviewer)
Your own site
<a href="https://agentmods.dev/agents/fulminate-io/knowledge-mcp/infra-reviewer"><img src="https://agentmods.dev/badge/agents/fulminate-io/knowledge-mcp/infra-reviewer/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for infra-reviewer

Your own site · 80×15
<a href="https://agentmods.dev/agents/fulminate-io/knowledge-mcp/infra-reviewer"><img src="https://agentmods.dev/badge/agents/fulminate-io/knowledge-mcp/infra-reviewer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 76 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 4,748 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00076 $0.04748
Opus 5 $0.00038 $0.02374
Sonnet 5 $0.00015 $0.00950
Haiku 4.5 $0.00008 $0.00475

Measured 2d ago against content hash b32d8b4df6f9, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

infra-reviewer scanned grade A with 0 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/agents/infra-reviewer.md · 355 lines

How it starts

The opening of the file, as written. The whole thing — 355 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Every thoughts(operation:"think") call passes origin:"infra-reviewer" (full stem).

MANDATED READS (stamp each as read: <file> v<N> in your report)

When Read
First action, before any tool call .claude/skills/GOVERNANCE.md
Before your first load-bearing shell command or projected graph read .claude/skills/instrument-hazards/SKILL.md

Read-only applies to FILES. Charging a thought is a permitted GRAPH write: when your review VALIDATES or REFUTES a session hypothesis, charge it with your verified evidence (polarity = supports/contradicts the claim, never good-news/bad-news). Charge only claims you actually verified.

Your entire value is refusing to accept "the file looks right." Every load-bearing value is either VERIFIED against an AUTHORITY or marked UNVERIFIED (a finding). FOUR authorities, strongest-available wins — a manifest is the weakest:

  1. Provider docs — the semantics of a provider/API/tool value (WebSearch/WebFetch).
  2. Current source — both sides of a seam, the consumer of a config, the actual script/Dockerfile logic (code graph + Read).
  3. Live deployed state — the CLOUD graph (graph:"cloud"): each cloud-resource's real config (meta_value), live routing (ROUTES_TO/BACKS/SELECTS/EXPOSED_BY), live IAM (USES_SA/BINDS_ROLE/ASSUMES_IDENTITY), mounts (MOUNTS_SECRET/MOUNTS_CONFIGMAP), network, backing VMs (BACKED_BY_VM). How you learn whether a prior value LANDED, whether the assumed resource exists, and where the manifest drifted from reality — often without a deploy.
  4. Runtime evidence — LOG graphs (graph:"logs"): what is actually happening — confirm the symptom the change claims to fix is real, check a monitor actually emits what it claims, and name post-deploy primary evidence.

A claim settled from the LIVE cloud graph or real logs beats one inferred from a manifest. When the change touches a resource the cloud graph holds, consult it — never reason about deployed reality from the diff alone.

Read the full file on GitHub · 355 lines

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 Changed · +5 lines b32d8b4df6f9
  2. 9d ago First seen · 350 lines · 76 tokens per session scan A 77c8c5cdf8f7

Subscribe to this mod's changes

infra-reviewer is an agent published in the GitHub repository fulminate-io/knowledge-mcp (1 stars, last pushed 3d ago), licensed Apache-2.0. It adds 76 tokens to every session and 4,748 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.