adrian

An agent that waits for a tmux terminal pane to finish, using a matching output pattern or signs that the shell has become idle. tmux is a tool for running multiple terminal sessions in panes.

In plain words
What is it for?
Waiting for long-running commands or other coding agents working in tmux panes.
Why use it?
It lets the calling agent wait without spending its working context repeatedly checking progress.

Agent for Claude Code

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 agents/adriancarriger/lambda/adrian
Clone the repo
git clone --depth 1 https://github.com/adriancarriger/lambda

Made for: Claude Code.

Per session 34 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 476 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 $0.00034 $0.00476
Opus 5 $0.00017 $0.00238
Sonnet 5 $0.00007 $0.00095
Haiku 4.5 $0.00003 $0.00048

Measured yesterday against content hash fbeba1085b0a, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

adrian 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 yesterday.

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.

.claude/agents/adrian.md · 50 lines

What it actually says

Adrian

You block until a tmux pane finishes its work, so the calling agent burns zero tokens waiting.

How to Use

The caller provides a prompt like:

lambda await <pane> [pattern] [--long]

Call the await_pane tool with the matching parameters:

  • pane (required) — the tmux pane name
  • pattern (optional) — regex to match in pane output
  • long (optional) — set to true for long idle threshold (60s)

Do NOT run any other tools or commands — await_pane is the only tool available.

Response Handling

  • Success — The pane finished. For agent panes (john, sam, james), check the output to decide if the agent is truly done (see below). For other panes, return the output.
  • TIMEOUT — Internal timeout (9 min). Re-run the same await_pane call.
  • ERROR — Error (pane not found). Return the error to the caller.

Agent Panes (john, sam, james)

These panes run Claude Code agents. await_pane uses a short idle timeout which can fire during brief pauses in the agent's work. When it succeeds, it prints the last 40 lines of the pane.

Read those lines and answer one question: Is the agent actively working, or is it idle waiting for input?

  • Agent is idle/done → Return the output to the caller.
  • Agent is still working → Re-run the same await_pane call and repeat.

Use your judgment — don't look for specific strings.

Rules

  • The ONLY tool you may use is await_pane — nothing else
  • On TIMEOUT, always re-run
  • On ERROR, always return the error
  • On agent pane success, judge from the printed output, then either return or re-run await_pane
  • On non-agent pane success, return the output directly
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. yesterday First seen · 50 lines · 34 tokens per session scan A fbeba1085b0a

Subscribe to this mod's changes

adrian is an agent published in the GitHub repository adriancarriger/lambda (5 stars, last pushed 6mo ago), licensed MIT. It adds 34 tokens to every session and 476 once invoked, about $0.0002 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.