release

release is a skill for Claude Code, Codex from eugene1g/agent-safehouse. It costs 57 tokens per session (3,255 once invoked), scanned A, original, Apache-2.0.

A release workflow for Agent Safehouse, covering version selection, changelog entries, GitHub releases, and the stable Homebrew package source.

In plain words
What is it for?
Use it to inspect commits, propose or set a version, update the changelog, and publish the GitHub and Homebrew release after confirmation.
Why use it?
It makes releases repeatable and gives the maintainer a dry run to review before publishing changes.

Skill for Claude CodeCodex

About the project

Agent Safehouse is a macOS sandboxing tool that restricts local AI coding agents to the files and integrations they are allowed to use. It is for developers who want to reduce the damage an agent can cause while preserving normal development workflows. Its catalogue entries help configure or use the sandbox with coding agents.

eugene1g/agent-safehouse · 2,050 stars · on GitHub

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/eugene1g/agent-safehouse/release
Any agent
npx skills add eugene1g/agent-safehouse --skill release
Clone the repo
git clone --depth 1 https://github.com/eugene1g/agent-safehouse

Made for: Claude Code, Codex.

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 release

README.md
[![agentmods](https://agentmods.dev/badge/skills/eugene1g/agent-safehouse/release.svg)](https://agentmods.dev/skills/eugene1g/agent-safehouse/release)
Your own site
<a href="https://agentmods.dev/skills/eugene1g/agent-safehouse/release"><img src="https://agentmods.dev/badge/skills/eugene1g/agent-safehouse/release.svg" alt="Measured on agentmods" height="20"></a>
Per session 57 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,255 The whole file, excluding the scripts and references it only reads on demand.
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 $0.00057 $0.03255
Opus 5 $0.00028 $0.01628
Sonnet 5 $0.00011 $0.00651
Haiku 4.5 $0.00006 $0.00326

Measured 5d ago against content hash fa90783b3249, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

release 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 5d 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/skills/release/SKILL.md · 217 lines

How it starts

The opening of the file, as written. The whole thing — 217 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Release

Use this skill when the user wants changelog text, release notes, or a CHANGELOG.md update for Agent Safehouse. This skill is also the canonical local release workflow for Agent Safehouse.

Historical changelog sections are append-only by default. Do not rewrite, normalize, or reformat older release entries to match a newer structure unless the user explicitly asks for a changelog migration or cleanup pass.

This skill is responsible for proposing the next release version from the last published release. The repo-root VERSION file is the canonical version source for Agent Safehouse. Write the selected version into VERSION and into the new CHANGELOG.md heading unless the user explicitly gives a different target version. Default to a stable release. Only choose a prerelease version such as -rc.N or -beta.N when the user explicitly asks for a prerelease. If the user explicitly gives the target version, honor it even if the default SemVer analysis would have chosen a different bump. In that case, note briefly in the dry run that the version came from an explicit user override.

Workflow

  1. Resolve the baseline release. If the user gives a starting version, use it. Otherwise:

    • Prefer the latest published stable GitHub release tag from gh release list --exclude-drafts --exclude-pre-releases.
    • Ignore draft releases.
    • Ignore prereleases such as -rc, -beta, and other hyphenated SemVer suffixes unless the user explicitly asks to prepare a prerelease.
    • If GitHub release metadata is unavailable, fall back to the latest stable SemVer git tag.
  2. Decide the next SemVer version from the release range. Start from the resolved baseline release and choose:

    • major for breaking changes, incompatible CLI behavior, removed capabilities, changed defaults that are likely to break existing setups, or sandbox/policy tightening that requires user action to restore a previously supported workflow.
    • minor for backward-compatible new features, new integrations, new profiles, new user-visible capabilities, or meaningful expansions of supported workflows.
    • patch for bug fixes, non-breaking sandbox/profile corrections, docs, tests, internal cleanup, and other maintenance-only changes.
    • If the range mixes features and fixes but nothing breaking, choose minor.
    • If the range is ambiguous and there is no clear breaking change or new feature, choose patch.
    • If you choose major, explicitly state why in the drafted release notes.
    • If the user explicitly asks for a prerelease, first choose the target stable version using the rules above, then append the requested prerelease suffix such as -rc.1 or -beta.1.
    • If the user explicitly asks for another prerelease in the same series, increment that series from the latest matching non-draft prerelease tag or release, for example 1.4.0-rc.1 to 1.4.0-rc.2.

Read the full file on GitHub · 217 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. 5d ago First seen · 217 lines · 57 tokens per session scan A fa90783b3249

Subscribe to this mod's changes

release is a skill published in the GitHub repository eugene1g/agent-safehouse (2,050 stars, last pushed today), licensed Apache-2.0. It adds 57 tokens to every session and 3,255 once invoked, about $0.0003 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.