sheena AGENTS.md

Repository instructions for Sheena, a Go library that runs an AI agent’s requested tool calls in a protected local sandbox. They describe the required Go version, coding style, tests, coverage, and access-control behavior.

In plain words
What is it for?
Use them when developing or reviewing Sheena code, especially when adding sandbox behavior, writing Go tests, checking race conditions, or maintaining coverage and benchmarks.
Why use it?
They give contributors one set of rules for making safe, compatible changes in the repository. They also define checks that changes must pass before they are accepted.

Instructions file for CodexOpenCode

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 instructions/goccy/sheena/agents-md
Clone the repo
git clone --depth 1 https://github.com/goccy/sheena

Made for: Codex, OpenCode.

Per session 1,152 This file is loaded in full into every session.
When invoked 1,152 The same file — it is already loaded in full.
Security scan A 0 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 $0.01152 $0.01152
Opus 5 $0.00576 $0.00576
Sonnet 5 $0.00230 $0.00230
Haiku 4.5 $0.00115 $0.00115

Measured 3d ago against content hash 096327e73e42, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

sheena AGENTS.md 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 3d 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.

AGENTS.md · 95 lines

How it starts

The opening of the file, as written. The whole thing — 95 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Guidance for AI agents and human contributors working in this repository. These rules are mandatory for every change. (A Claude-specific superset lives in CLAUDE.md; this file is the tool-neutral version.)

What Sheena is

A lightweight, in-process sandbox for safely executing the tool calls an AI agent requests, on the same host as the agent. See README.md for the API and DESIGN.md for the full design.

Go version & idioms

  • Target Go 1.25.0. Any feature up to and including Go 1.25 may be used (generics, os.Root, io/fs.ReadLinkFS, range-over-func, etc.).
  • Prefer modern idioms. Use any, never interface{}. Use generics where they remove duplication or improve type safety.
  • Keep go.mod's go 1.25.0 directive and the golangci-lint config in sync.

Testing & coverage

  • Every package MUST maintain ≥90% statement coverage. Add tests until the bar is met; never lower it.
  • make test (go test -race ./..., with the wasm interpreters run separately without -race) and make cover (per-package ≥90% gate) MUST pass.
  • Access-control behavior is documented and asserted with example tests (Example… + // Output:) — path traversal blocked, network denied by default, overlay never writes to the host.
  • Keep the dedicated suites: security/threat-model (security_test.go), escape attempts (traversal, symlink escape, SSRF, redirect, output/command/memory caps, timeout), and the differential golden tests that check commands against real coreutils (internal/difftest).

Benchmarks

Sheena claims to be lightweight; that claim stays measured. Keep Benchmark… for sandbox construction/mass startup (report allocs/op and B/op), per-call ExecBash overhead, and parallel throughput (b.RunParallel). Run with make bench; compare with benchstat when touching hot paths and do not regress startup or allocations without a stated reason.

Linting

golangci-lint MUST pass with zero issues (make lint). Fix the underlying issue rather than disabling a linter or adding a nolint directive; if a suppression is truly warranted, justify it in the diff.

Read the full file on GitHub · 95 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. 3d ago First seen · 95 lines · 1,152 tokens per session scan A 096327e73e42

Subscribe to this mod's changes

sheena AGENTS.md is an instructions file published in the GitHub repository goccy/sheena (22 stars, last pushed 1mo ago), licensed MIT. It adds 1,152 tokens to every session, about $0.0058 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 2026-08-30.