Tenuo skill
A skill that teaches an agent to add authorization at the point of effect: warrants limit tools and arguments, tests check denial.
- 106 ★ on the repository
- Last commit
- Apache-2.0
Minimal access
Why this level
- The skill edits your project authorization code, so review its changes
- A wrongly placed check creates a false sense of protection
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/tenuoDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill edits your project authorization code, so review its changes
- A wrongly placed check creates a false sense of protection
This is third-party code. Review the repository files before installing.
What it does
Tenuo gives each task only the authority it needs. A warrant is a signed permission to call specific tools with constraints on arguments and time, like a prepaid card for one task. Rights can only shrink when handed on, are checked where the action runs, and sensitive steps require a signed human approval. Libraries exist for Python (`pip install tenuo`), TypeScript (beta) and Rust, and the authorizer deploys in-process, as a sidecar or at a gateway. Installing the `tenuo-agent-authorization` skill helps a coding agent place authorization at the effect boundary, use the SDK version from your project and write tests for allow and deny before the effect. The skill distinguishes an in-process guardrail from an independently verifiable boundary and names the remaining bypasses. The repository also holds other skills: `tenuo-audit`, `tenuo-denial-triage` and `tenuo-warrant`.
Who it is for. Developers of agents and platforms who give models access to tools and want to limit it by task rather than by the whole account.
Good fit when
- An agent delegates to subagents and rights must not grow when handed on.
- A dangerous tool, such as cluster scaling, must be constrained by arguments.
- Authorization denials are analysed after an agent failure.
- An audit must show signed evidence of allowed and denied actions.
Not a fit when
- Account-level permissions without per-task limits are enough.
- A ready cloud service is needed: Tenuo Cloud is in open beta and the core installs as a library.
- The project is not in Python, TypeScript or Rust.
Example request
Use tenuo-agent-authorization: protect the scale_cluster function with a warrant constrained by cluster and add a denial test.Limitations
The skill helps embed the library, and the guarantee depends on where the check happens: inside the agent process it is weaker than at an independent boundary, which the skill reports separately. The TypeScript SDK is in beta and needs Node.js 20 or newer. The managed Tenuo Cloud platform is in open beta and its platform part is not in the repository. The agent must not reimplement the cryptography and protocol. The repository was last updated on 6 November 2026 and uses the Apache-2.0 licence, with the core marked Production/Stable v0.2 in the README.
How to disable. Delete the `tenuo-agent-authorization` folder from the agent skills directory. Remove the library with the normal package manager command and drop authorization checks added to the project through Git history.
FAQ
What is a warrant?
A signed permission to call specific tools with limits on arguments and time; rights only shrink when handed on.
Which skills does the repository include?
`tenuo-agent-authorization`, `tenuo-audit`, `tenuo-denial-triage` and `tenuo-warrant`.
Where is the check performed?
In-process, in a sidecar or at a gateway; the skill distinguishes an in-process guardrail from an independently enforced boundary.
Which languages are supported?
Python via `pip install tenuo`, TypeScript in beta via `npm i @tenuo/core@beta` and Rust via `cargo add tenuo --features sdk`.
Can it be tried without installing?
Yes, there is a Colab notebook, a warrant explorer and a Docker delegation-chain demo.
Discussion
No questions or reviews yet.
Related
SkillsVibeSec
VibeSec-Skill
A secure coding skill: the agent looks at code from a bug hunter's perspective and prevents common web vulnerabilities
A security scanner for agent skills: finds prompt injection, data exfiltration and malicious code before you install
A code security audit skill by Cloudflare: the agent runs recon, coverage-led hunting and independent verification of findings, then produces a structured repor
Rust OS-level sandbox and MCP server for agent skills: process isolation, install-time scanning and self-evolution
A taint analysis engine for Java and Kotlin with agent skills: scans code, triages findings and writes rules.
A skill and local program for agent safety: inventory, skill admission checks, human-approved grants and signed receipts.
README in short
The README explains the warrant model, shows protecting one function and moving to delegation, and the agent section describes installing the skill.