What is fleet CLAUDE.md?

Repository instructions for Fleet, a terminal interface for managing multiple coding-agent sessions. They cover its Go technology stack, build and test commands, commit format, and changelog rules.

Building, running, formatting, testing, linting, and documenting Fleet, including writing correctly formatted commits and changelog entries.

They provide the project context and contribution rules needed to change Fleet without breaking its build, tests, versioning, or release notes.

fleet CLAUDE.md — Instructions file for coding agents by brizzai: Repository instructions for Fleet, a terminal interface for managing…

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/brizzai/fleet/claude-md

Clone the repo

git clone --depth 1 https://github.com/brizzai/fleet

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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/brizzai/fleet/claude-md.svg)](https://agentmods.dev/instructions/brizzai/fleet/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/brizzai/fleet/claude-md"><img src="https://agentmods.dev/badge/instructions/brizzai/fleet/claude-md.svg" alt="Measured on agentmods" height="20"></a>

Per session 35,266 This file is loaded in full into every session.

When invoked 35,266 The same file — it is already loaded in full.

Security scan B 2 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.

Model	Per session	Once invoked
Fable 5	$0.35266	$0.35266
Opus 5	$0.17633	$0.17633
Sonnet 5	$0.07053	$0.07053
Haiku 4.5	$0.03527	$0.03527

Measured 3d ago against content hash 6c407ffab47c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

fleet CLAUDE.md scanned grade B with 2 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.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

- Hooks auto-installed into `~/.claude/settings.json` on TUI launch

Runs shell commandslowCapability

Expected in a hook, worth knowing in a rule or an instructions file.

- OpenCode status: driven entirely by a generated TS plugin (no pane scraping); same agent-neutral `fleet hook-handler` pipeline. Unlike Claude/Codex declarative hook JSON, OpenCode's hook mechanism is a JS/TS plugin, so

CLAUDE.md · 270 lines

How it starts

The opening of the file, as written. The whole thing — 270 lines — stays where its author put it; the contents beside it link to each section on GitHub.

fleet

This file provides context for AI-assisted development with Claude Code.

TUI tool for managing multiple Claude Code and OpenAI Codex sessions in parallel using tmux.

Tech Stack

Go 1.26+, Bubble Tea v2 + Lip Gloss v2 (charm.land/.../v2), charmbracelet/x/vt (drawer's live terminal), tmux, SQLite (WAL mode)

Build

make build    # build to build/fleet
make run      # go run
make test     # go test -race
make fmt      # go fmt
make lint     # golangci-lint
make coverage # test coverage report

Commit Convention

Use conventional commits. Version is auto-computed from commits:

fix: ... → patch (v0.1.0 → v0.1.1)
feat: ... → minor (v0.1.0 → v0.2.0)
feat!: ... or BREAKING CHANGE: ... → major (v0.1.0 → v1.0.0)
chore:, docs:, refactor:, test:, style: → patch
Scopes are optional: fix(hooks): ..., feat(ui): ...

Changelog

Each PR adds a fragment file: changelog/unreleased/<name>.md with type: frontmatter (added/improved/fixed/changed/removed)
Keep fragments concise: 1–2 sentences, lead with the user-facing change. No implementation detail, marketing tone, or exhaustive sub-points — a user skimming the release notes should grasp it at a glance. e.g. "Status updates now respond in ~150ms instead of up to 2s."
Voice — fun via honest specificity, not hype. Fragments render in-app (Ctrl+K → "Release Notes", which formats `code` + **bold**), so write for a reader: open each bullet with a bold 2–4 word headline so a release skims in one pass; write in second person ("You can now…", "Sessions now…"); name the annoyance you killed ("…instead of a cryptic 'Operation not permitted'"); numbers over adjectives ("~150ms instead of 2s"); code-format keys/commands/paths (Ctrl+K, ~/Documents); one idea per bullet; ban hype words (seamless, powerful, delightful, blazing). Full guide + examples: changelog/README.md.
Add highlight: true to a fragment's frontmatter to also surface it in the in-app What's New reel. The item still renders in its normal type section; the flag also copies it into a ### Highlights block that leads the release (emitted by prepare-release.yml). Reserve for changes a user would be glad to discover — most fragments are not highlights.
Never decide highlight on your own — always ask. When you write a fragment, ask the user (via the AskUserQuestion tool) whether it should be a highlight, and give a recommendation with your reasoning. A highlight leads the release in the What's New reel, so it's an editorial call about what deserves the user's attention — not a property of the diff, and not something you can infer from how big the change felt to implement. Default recommendation is no: most fragments are not highlights. Ask even when you're confident, and ask on the way in (when writing the fragment), not after the PR is up.
CI checks for fragment presence; comment /no-changelog to skip
At release time, fragments are merged into CHANGELOG.md and deleted

Read the full file on GitHub · 270 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.

3d ago First seen · 270 lines · 35,266 tokens per session scan B 6c407ffab47c

Subscribe to this mod's changes

fleet CLAUDE.md is an instructions file published in the GitHub repository brizzai/fleet (52 stars, last pushed 7d ago), licensed Apache-2.0. It adds 35,266 tokens to every session, about $0.1763 per session on Opus 5. A static security scan graded it B with 2 findings (reads agent configuration directories, runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

fleet CLAUDE.md