repository-manager-fleet-scale-operations

repository-manager-fleet-scale-operations is a skill for Claude Code, Codex from Knuckles-Team/repository-manager. It costs 141 tokens per session (2,205 once invoked), scanned A, original, MIT.

A workflow for coordinating development across many repositories and concurrent work lanes. A worktree is a separate working copy of a Git repository, and a merge queue is an ordered list of changes waiting to be integrated.

In plain words
What is it for?
Use it to create worktrees in bulk, classify and prune stale ones, drain merge queues across repositories, and set a safe concurrency level for heavy jobs.
Why use it?
Large-scale parallel work can exhaust disk, memory, or swap space and can accidentally remove work that is still active. This workflow helps audit, schedule, and clean up that work safely.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Use it to create worktrees in bulk, classify and prune stale ones, drain merge queues across repositories, and set a safe concurrency level for heavy jobs.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/knuckles-team/repository-manager/repository-manager-fleet-scale-operations
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.

Any agent
npx skills add Knuckles-Team/repository-manager --skill repository-manager-fleet-scale-operations
Clone the repo
git clone --depth 1 https://github.com/Knuckles-Team/repository-manager

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 repository-manager-fleet-scale-operations

README.md
[![agentmods](https://agentmods.dev/badge/skills/knuckles-team/repository-manager/repository-manager-fleet-scale-operations/github.svg)](https://agentmods.dev/skills/knuckles-team/repository-manager/repository-manager-fleet-scale-operations)
Your own site
<a href="https://agentmods.dev/skills/knuckles-team/repository-manager/repository-manager-fleet-scale-operations"><img src="https://agentmods.dev/badge/skills/knuckles-team/repository-manager/repository-manager-fleet-scale-operations/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 repository-manager-fleet-scale-operations

Your own site · 80×15
<a href="https://agentmods.dev/skills/knuckles-team/repository-manager/repository-manager-fleet-scale-operations"><img src="https://agentmods.dev/badge/skills/knuckles-team/repository-manager/repository-manager-fleet-scale-operations.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 141 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,205 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 $0.00141 $0.02205
Opus 5 $0.00071 $0.01103
Sonnet 5 $0.00028 $0.00441
Haiku 4.5 $0.00014 $0.00220

Measured 10d ago against content hash 3b81bd07ac20, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

repository-manager-fleet-scale-operations 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 10d 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.

repository_manager/skills/repository-manager-fleet-scale-operations/SKILL.md · 195 lines

How it starts

The opening of the file, as written. The whole thing — 195 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Repository Manager — Fleet-Scale Operations

Operating ~234 repositories with dozens of concurrent lanes. The failure modes at this scale are different in kind from a single lane's: you do not lose work to a merge conflict, you lose it to a machine that ran out of memory during a git add, or to a mass prune that deleted a branch someone was still holding.

When to use

  • Launching a wave of lanes across several repositories.
  • Reclaiming accumulated worktrees and branches.
  • Draining more than one repository's merge queue.
  • Deciding how many heavy jobs may run concurrently.

When NOT to use

  • One lane's lifecycle → repository-manager-lane-lifecycle.
  • Landing/reconciling one branch → repository-manager-merge-and-reconcile.

★ The binding constraint is disk I/O and swap — not the agent cap

The instinct is to size a wave by how many agents you may run. That is the wrong resource. Measured at peak on this host: load ~26 on 24 cores, swap 100% exhausted, 14 concurrent uv sync against a 112 GB cache — and a plain git add died with Bus error. That is data loss caused by scheduling, not by any lane doing anything wrong.

Cap HEAVY lanes at ~3–4 concurrently, independent of total lane count. Heavy means anything that does bulk I/O or bulk compilation:

Heavy (cap 3–4) Light (costs almost nothing)
uv sync / any dependency resolution editing files
cargo build / cargo check reading, searching, KG queries
pre-commit run --all-files git commit, git status
a full pytest sweep --lane doctor, --merge-queue status
container image builds writing docs/skills

Light lanes are nearly free; run as many as you have work for. The mistake is treating all lanes as equivalent and sizing the whole wave by the heaviest one — or worse, by the agent cap, which measures nothing physical.

Two corollaries:

  • Serialize the LEASE-class heavies. uv sync/uv lock, pre-commit --all-files, and any venv swap are already LEASE-class because of this. Run them through the lease and let contenders defer: agent-utilities lane lease --resource dependency-lock --operation relock -- uv lock. Exit 75 = another lane holds it.
  • Prune build artifacts at verify time, not at the end. A shared cargo target dir has already hit ENOSPC mid-wave. Each lane builds into ./target-isolated and removes it when its gate is green.

Read the full file on GitHub · 195 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. 10d ago First seen · 195 lines · 141 tokens per session scan A 3b81bd07ac20

Subscribe to this mod's changes

repository-manager-fleet-scale-operations is a skill published in the GitHub repository Knuckles-Team/repository-manager (2 stars, last pushed 13d ago), licensed MIT. It adds 141 tokens to every session and 2,205 once invoked, about $0.0007 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 2026-08-31.