defer

defer is a command for coding agents from Vitalini/ebb-ai. It costs 168 tokens per session (1,893 once invoked), scanned A, original, Apache-2.0.

A command that sends non-urgent tasks to a scheduler, which chooses a lower-carbon electricity period before a deadline.

In plain words
What is it for?
Use it to defer coding or other agent tasks to a specified or inferred deadline. The task is sent to the ebb-ai scheduler for later execution.
Why use it?
It lets work happen later when the electricity supply is cleaner, instead of running it immediately. It turns everyday phrases such as “by tomorrow” or “overnight” into a scheduled task.

Command

Part of the ebb-ai plugin — 1 skill, 8 commands, 1 MCP server shipped together

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 commands/vitalini/ebb-ai/defer
Clone the repo
git clone --depth 1 https://github.com/Vitalini/ebb-ai

Or install ebb-ai, the plugin that ships this one along with the rest of its 1 skill, 8 commands, 1 MCP server.

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 defer

README.md
[![agentmods](https://agentmods.dev/badge/commands/vitalini/ebb-ai/defer.svg)](https://agentmods.dev/commands/vitalini/ebb-ai/defer)
Your own site
<a href="https://agentmods.dev/commands/vitalini/ebb-ai/defer"><img src="https://agentmods.dev/badge/commands/vitalini/ebb-ai/defer.svg" alt="Measured on agentmods" height="20"></a>
Per session 168 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,893 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 $0.00168 $0.01893
Opus 5 $0.00084 $0.00946
Sonnet 5 $0.00034 $0.00379
Haiku 4.5 $0.00017 $0.00189

Measured 5d ago against content hash 69eb8f501b4f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

defer 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 5d 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.

packages/claude-code-plugin/commands/defer.md · 133 lines

How it starts

The opening of the file, as written. The whole thing — 133 lines — stays where its author put it; the contents beside it link to each section on GitHub.

The user wants to defer the following task to a cleaner-grid window using the ebb-ai MCP server. This is the central UX of the ebb-ai Claude Code plugin — take a "do it later" request and route it to a real scheduler.

When to invoke automatically (without the user explicitly typing /ebb-ai:defer)

Any of these signals means the user has implicitly opted into deferral. Convert their wording to an ISO-8601 deadline and call schedule_task:

User phrase Implied deadline
"by tomorrow" tomorrow, end of working day (18:00)
"tonight" / "overnight" tomorrow, 06:00 local
"by EOD" today, 18:00 local
"this week" Friday, 18:00 local
"next week" next Friday, 18:00 local
"when you have a moment" +24 h
"remind me to X" +24 h, unless a time is named
"no rush" / "not urgent" +48 h
"schedule this for X" parse X explicitly
"queue this up" +12 h

If the user does NOT signal "later" (e.g. "summarize this", "write a function", "what does X do"), do not auto-defer. Run inline.

If the user explicitly typed /ebb-ai:defer ..., follow the explicit args. The auto-invocation above is for the unprompted case only.

Arguments

$ARGUMENTS

Parse $ARGUMENTS into:

  • Prompt — everything before the first -- flag, or the whole argument string if no flags are present. Required. If empty, ask the user what task they want to defer.
  • --by <duration> — deadline. Accepts 2h, 30m, 24h, 3d, or an ISO-8601 timestamp. Default: 24h from now. Must resolve to the future.
  • --budget <grams> — optional hard cap on grams CO2-equivalent. If the scheduler can't find a window that meets it, the task fails fast with a CarbonBudgetExceededError.
  • --region <zone> — optional Electricity Maps zone code (e.g. US-CAL-CISO, GB, FR). Default: omit the field — the server resolves its own default (EBB_DEFAULT_REGION if configured, otherwise a guess from the host machine's timezone, falling back to GB). The region line in the tool response tells you what it picked; relay that to the user.
  • --model <name> — optional vendor model hint (e.g. claude-sonnet-4-6). Required when the deferred task should actually dispatch via ebb tick — without a model the server falls back to its default (EBB_DEFAULT_MODEL, else claude-sonnet-4-6).
  • --output <abs-path> — optional absolute file path. When the task completes, ebb-ai writes { taskId, result, receipt } as JSON to this path. Useful if the user wants the result to land in an inbox folder (so tail -f or a file-watcher surfaces it) rather than polling /ebb-ai:check by hand. Recommend ~/.ebb-ai/inbox/<task_id>.json when the user is unsure where to put it.
  • --provider <anthropic|openai> — optional. Default anthropic.

Read the full file on GitHub · 133 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. 5d ago First seen · 133 lines · 168 tokens per session scan A 69eb8f501b4f

Subscribe to this mod's changes

defer is a command published in the GitHub repository Vitalini/ebb-ai (1 stars, last pushed 24d ago), licensed Apache-2.0. It adds 168 tokens to every session and 1,893 once invoked, about $0.0008 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.