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.
npx skills add UnboundCompute/security-agent-skills --skill auditing-kms-key-policy-and-envelope-encryptiongit clone --depth 1 https://github.com/UnboundCompute/security-agent-skillsWrote 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.
[](https://agentmods.dev/skills/unboundcompute/security-agent-skills/auditing-kms-key-policy-and-envelope-encryption)<a href="https://agentmods.dev/skills/unboundcompute/security-agent-skills/auditing-kms-key-policy-and-envelope-encryption"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/auditing-kms-key-policy-and-envelope-encryption/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.
<a href="https://agentmods.dev/skills/unboundcompute/security-agent-skills/auditing-kms-key-policy-and-envelope-encryption"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/auditing-kms-key-policy-and-envelope-encryption.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00171 | $0.01906 |
| Opus 5 | $0.00086 | $0.00953 |
| Sonnet 5 | $0.00034 | $0.00381 |
| Haiku 4.5 | $0.00017 | $0.00191 |
Grade A, and why
auditing-kms-key-policy-and-envelope-encryption 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 11d 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.
How it starts
The opening of the file, as written. The whole thing — 134 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Auditing KMS key policy and envelope encryption: who can actually decrypt
Envelope encryption protects data with a data key and protects the data key with a managed master key, so the real access boundary is the master key's policy plus whatever encryption context constrains the data key. That indirection hides over-broad access. A key policy or grant can admit a principal who should never decrypt this data; a wildcard key resource in an identity policy can cover keys the author never considered; and an encryption context, meant to bind a data key to a specific scope, does nothing if the decrypt path does not enforce it. Cross-account grants widen the set further. The payload can be perfectly encrypted while the decrypt set quietly includes the wrong principals. You audit these by computing who can decrypt and comparing that to who should read the data.
When to use
- Data is protected by a managed key through envelope encryption, with a data key under a master key.
- Decrypt access is governed by a key policy, key grants, and identity policies that may use wildcards.
- An encryption context is used to scope a data key, and you need to confirm it is enforced on decrypt.
Scope check
Audit key access only in accounts and key stores you own or are authorized to assess. Confirming a decrypt recovers real plaintext, so stay inside the authorized keys and treat recovered data as sensitive within scope, never exfiltrating it. If you can't name the authorization, stop.
The loop
-
Establish the data's intended readers first. For the data a key protects, name the principals that legitimately need to decrypt it. This is the false-positive killer: a key policy admitting exactly those readers is correct however sensitive the key is. Name the intended readers, then compute the decrypt set against them.
-
Read the key policy and grants for who they admit. The master key's policy and any grants name the principals allowed to decrypt. Determine the true admitted set, noting broad principals (an account-wide grant, a wildcard) and any condition intended to scope it. Grants are easy to miss because they live apart from the policy document.
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.
- 11d ago First seen · 134 lines · 171 tokens per session scan A 7a4347b7f847
auditing-kms-key-policy-and-envelope-encryption is a skill published in the GitHub repository UnboundCompute/security-agent-skills (5 stars, last pushed 3d ago), licensed MIT. It adds 171 tokens to every session and 1,906 once invoked, about $0.0009 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.
Other skills, from other repositories
prodcheck-review
Review this codebase against the prodcheck pre-production checklists — security, performance, scale, integrations and post-launch readiness. Use when asked to check whether a project is ready to ship, to audit an area before launch, or to work through a specific checklist. Produces evidence with file:line citations…
integrate-arcjet-guard-genkit
Integrate Arcjet security into a Genkit JS agent using @arcjet/guard — wrap ai.defineTool, put guardMiddleware on generate({ use }) for unwrapped / MCP / filesystem tools, and read a caller-owned id from generate({ context }). Use when asked to add Arcjet to genkit, rate limit its tools, screen inbound messages, or…
integrate-arcjet-guard-langgraph
Integrate Arcjet security into a LangGraph Graph API agent using @arcjet/guard — wrap tool() / StructuredTool, wrap ToolNode for unwrapped MCP tools, and read threadid for correlation. Use when asked to add Arcjet to a LangGraph StateGraph / ToolNode agent, rate limit its tools, screen inbound messages, or block…
integrate-arcjet-guard-tanstack-ai
Integrate Arcjet security into a TanStack AI chat() app using @arcjet/guard — put guardMiddleware first on chat({ middleware }) so onBeforeToolCall gates tools, and read a caller-owned id from chat({ context }). Use when asked to add Arcjet to TanStack AI, rate limit its tools, screen inbound messages, or block prompt…
integrate-arcjet-guard-claude-managed-agents
Integrate Arcjet security into Claude Managed Agents (hosted REST+SSE, beta managed-agents-2026-04-01) using @arcjet/guard — screen user.message / initialevents before sessions.events.send, and gate custom tools on agent.customtooluse. Use when asked to add Arcjet to Claude Managed Agents, rate limit custom tools, or…
linear-claude-skill
Manage Linear issues, projects, and teams.