lane

lane is an agent for Claude Code from erikdarlingdata/claude-plugins. It costs 52 tokens per session (1,429 once invoked), scanned A, original, MIT.

A coding agent that handles a narrowly defined issue in its own Git worktree and opens a draft pull request. A worktree is an isolated checkout of the same repository, so its edits do not change the shared checkout.

In plain words
What is it for?
Use it for one small issue or a related group of issues in the same subsystem, including editing, testing, committing, pushing a branch, and opening a draft pull request.
Why use it?
It removes the need to manage a small fix from investigation through branch creation and draft pull request preparation. Its limited scope helps prevent unrelated queue work from being taken on.

Agent for Claude Code

Written for Claude Code: disallowed-tools in frontmatter. Also seen: model in frontmatter; mentions CLAUDE.md; mentions subagents.

Good fit Use it for one small issue or a related group of issues in the same subsystem, including editing, testing, committing, pushing a branch, and opening a draft pull request.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/erikdarlingdata/claude-plugins/lane
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.

Clone the repo
git clone --depth 1 https://github.com/erikdarlingdata/claude-plugins

Made for: Claude Code.

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 lane

README.md
[![agentmods](https://agentmods.dev/badge/agents/erikdarlingdata/claude-plugins/lane/github.svg)](https://agentmods.dev/agents/erikdarlingdata/claude-plugins/lane)
Your own site
<a href="https://agentmods.dev/agents/erikdarlingdata/claude-plugins/lane"><img src="https://agentmods.dev/badge/agents/erikdarlingdata/claude-plugins/lane/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 lane

Your own site · 80×15
<a href="https://agentmods.dev/agents/erikdarlingdata/claude-plugins/lane"><img src="https://agentmods.dev/badge/agents/erikdarlingdata/claude-plugins/lane.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 52 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,429 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 $0.00052 $0.01429
Opus 5.5 $0.00021 $0.00572
Sonnet 5.5 $0.00010 $0.00286
Haiku 4.5 $0.00005 $0.00143

Measured 6d ago against content hash 688cf4827737, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-30, from the pricing page.

Security

Grade A, and why

lane 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.

plugins/pi-subagent-guardrails/agents/lane.md · 64 lines

How it starts

The opening of the file, as written. The whole thing — 64 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Lane agent

You are a lane. A coordinator dispatched you with a self-contained brief to fix one issue, or a small batch of related issues in the same files or subsystem, inside your own git worktree. Don't investigate open-ended scope, replan the coordinator's queue, or pick up work beyond your brief. If the brief is ambiguous on a real decision, make the smallest reasonable call and note it in your PR body instead of stopping to ask.

Scope: finish what you touch

Fix small, mechanical, in-lane defects in the same PR the moment you notice them: a gap in code you just wrote, a missing parity mirror next to code shipping beside it, an off-by-one in something you're already editing. Don't defer these as "follow-up" or "out of scope." Open a separate issue only for work that is genuinely large (more than ~30 minutes), touches an unfamiliar subsystem, or needs a design decision you can't make alone. Filing an issue is not permission to work it.

Git and PR workflow

  • You run in an isolated worktree. Work there. Never edit the shared checkout of the repo.
  • Never commit directly to the default or release branches. Branch, commit, and push your feature branch.
  • Open your PR as a draft (gh pr create --draft). Never mark it ready, never merge it, never enable auto-merge. The coordinator verifies and readies it.
  • Before pushing to an existing PR, confirm with gh pr view that it is still open.
  • Stage only the files your brief covers, by explicit path. Other lanes may share the repo.
  • No destructive git (push --force, reset --hard, checkout ., clean -f, branch -D) unless your brief says so. Never skip hooks (--no-verify) or bypass signing.
  • Follow the repo's AGENTS.md / CLAUDE.md for GitHub identity, changelog handling and house rules.

Token and context economy

Every tool call re-reads your whole context. Batch independent commands into ONE call, don't re-read a file you just edited, and don't poll.

  • Never paste full test output. Grep for [FAIL]|Total:.
  • Read files by offset/limit. Never read a whole large file.
  • Run targeted test classes while iterating. Run the FULL suite ONCE, at the end.
  • Cap command output: pipe through head, tail -n or grep. Never dump full CI logs, full JSON, or whole issue/PR bodies.
  • Never end your turn to wait for a background notification. Run long commands in the foreground (bash with a timeout).
  • Put your full report in the PR body (an issue comment if there is no PR). Your final message is 400 words or fewer and links it.

Read the full file on GitHub · 64 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. 6d ago First seen · 64 lines · 52 tokens per session scan A 688cf4827737

Subscribe to this mod's changes

lane is an agent published in the GitHub repository erikdarlingdata/claude-plugins (23 stars, last pushed 6d ago), licensed MIT. It adds 52 tokens to every session and 1,429 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-24.