agent-glovebox: Command for GitHub Copilot

.github/prompts/agent-publish-manifest.md

agent-publish-manifest is a command for GitHub Copilot from AlexanderMattTurner/agent-glovebox. It costs 0 tokens per session (699 once invoked), scanned A, original, Apache-2.0.

A controlled way to request selected GitHub publishing actions without giving the coding agent a GitHub write credential in its shell. GitHub is a service for hosting code, pull requests, and issue discussions.

In plain words
What is it for?
Use it to prepare a manifest requesting a pull request, issue update, comments, labels, or pull-request closures after making commits locally.
Why use it?
It separates local code changes from remote publishing, reducing the chance that an agent can push code or alter issues and pull requests outside the declared list.

Command for GitHub Copilot

Written for GitHub Copilot: a Copilot chat mode or prompt.

This is AlexanderMattTurner/agent-glovebox's own configuration. It tells GitHub Copilot how to work on agent-glovebox 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 agent-glovebox configures →

Reuse

Borrowing it

Nothing to install: this file belongs to AlexanderMattTurner/agent-glovebox. 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/AlexanderMattTurner/agent-glovebox/main/.github/prompts/agent-publish-manifest.md
Clone the repo
git clone --depth 1 https://github.com/AlexanderMattTurner/agent-glovebox

Made for: GitHub Copilot.

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 agent-publish-manifest

README.md
[![agentmods](https://agentmods.dev/badge/commands/alexandermattturner/agent-glovebox/agent-publish-manifest/github.svg)](https://agentmods.dev/commands/alexandermattturner/agent-glovebox/agent-publish-manifest)
Your own site
<a href="https://agentmods.dev/commands/alexandermattturner/agent-glovebox/agent-publish-manifest"><img src="https://agentmods.dev/badge/commands/alexandermattturner/agent-glovebox/agent-publish-manifest/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 agent-publish-manifest

Your own site · 80×15
<a href="https://agentmods.dev/commands/alexandermattturner/agent-glovebox/agent-publish-manifest"><img src="https://agentmods.dev/badge/commands/alexandermattturner/agent-glovebox/agent-publish-manifest.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 699 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 $0.00000 $0.00699
Opus 5 $0.00000 $0.00349
Sonnet 5 $0.00000 $0.00140
Haiku 4.5 $0.00000 $0.00070

Measured 6d ago against content hash 16959f6525e2, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

agent-publish-manifest 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 6d 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.

.github/prompts/agent-publish-manifest.md · 41 lines

What it actually says

You hold no GitHub credential in this run. git push, gh pr create, gh pr close and gh issue all fail, because the workflow keeps the write token out of your shell.

Commit your work locally as usual — git checkout -b, git add, git commit need no credential — then declare what you want published. The workflow's publish step performs that list under its own token. It is a fixed list of actions, and the workflow pins each one's target: which base branch a pull request opens against, which tracking issue the issue verb rewrites, which pull requests you may close, and which issues you may comment on. .github/scripts/agent_publish.py is the whole list.

Write one JSON object to $RUNNER_TEMP/agent-publish/manifest.json. Create the directory first. Every key is optional; an unknown key fails the step.

{
  "pull_request": {
    "title": "fix(security): weekly vulnerability remediation",
    "body_file": "/path/to/pr-body.md",
    "labels": ["security-scan"]
  },
  "close_pull_requests": [412, 413],
  "issue": {
    "title": "the scanner reports a new advisory with no upstream fix",
    "body_file": "/path/to/issue-body.md",
    "labels": ["security-scan"]
  },
  "issue_comments": [{ "number": 907, "body_file": "/path/to/comment.md" }]
}

Shipping nothing is the whole manifest instead, never a key beside the others:

{ "nothing_to_publish": "every alert was already covered by an open triage PR" }

What each key does:

  • pull_request — pushes the branch you are on and opens its pull request, or rewrites the title and body of the open one that branch already has. Create a branch first: the step refuses to push when HEAD is still the base branch. Commit everything too — it refuses a worktree with uncommitted changes to tracked files, because no pull request would carry them.
  • close_pull_requests — closes each number with the comment Superseded by #<your PR>. Needs pull_request in the same manifest. The step refuses a number whose head branch is outside what this job may subsume.
  • issue — rewrites this job's own tracking issue, or opens it. The workflow names that issue with a hidden marker, and the step adds the marker itself; you cannot select a different issue. labels apply only when it opens a new one.
  • issue_comments — one comment per entry. The step refuses a number outside the set the workflow staged for this run.
  • nothing_to_publish — your reason for shipping nothing. It must be the manifest's only key, and you must have made no commit.

Each body_file holds the finished text, written by you, under $RUNNER_TEMP or the workspace. The step refuses an empty file and a path outside those two directories.

Any commit you make must be published by a pull_request in this manifest. A commit with no manifest, a commit beside nothing_to_publish, and a commit whose manifest declares only an issue are all errors — each would drop your work, so the step reds instead.

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. 6d ago First seen · 41 lines · 0 tokens per session scan A 16959f6525e2

Subscribe to this mod's changes

agent-publish-manifest is a command published in the GitHub repository AlexanderMattTurner/agent-glovebox (63 stars, last pushed today), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 699 tokens. 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 2026-09-06.