flow-selector

flow-selector is a skill for Claude Code, Codex from smicolon/ai-kit. It costs 49 tokens per session (1,209 once invoked), scanned A, original, MIT.

A choice step that shows two or more possible task flows and asks the user to select one.

In plain words
What is it for?
Use it after multiple candidate flows have been found and the user must choose the one to follow.
Why use it?
It prevents implementation from proceeding with the wrong interpretation when a request can be handled in several ways.

Skill for Claude CodeCodex

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 skills/smicolon/ai-kit/flow-selector
Any agent
npx skills add smicolon/ai-kit --skill flow-selector
Clone the repo
git clone --depth 1 https://github.com/smicolon/ai-kit

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 flow-selector

README.md
[![agentmods](https://agentmods.dev/badge/skills/smicolon/ai-kit/flow-selector.svg)](https://agentmods.dev/skills/smicolon/ai-kit/flow-selector)
Your own site
<a href="https://agentmods.dev/skills/smicolon/ai-kit/flow-selector"><img src="https://agentmods.dev/badge/skills/smicolon/ai-kit/flow-selector.svg" alt="Measured on agentmods" height="20"></a>
Per session 49 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,209 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin unknown 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 $0.00049 $0.01209
Opus 5 $0.00024 $0.00605
Sonnet 5 $0.00010 $0.00242
Haiku 4.5 $0.00005 $0.00121

Measured 4d ago against content hash e7a3124a5f5e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

flow-selector 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 4d 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.

packs/clarify/skills/flow-selector/SKILL.md · 125 lines

How it starts

The opening of the file, as written. The whole thing — 125 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Flow Selector

Present candidate flows produced by flow-detector and let the user pick one. Single decision point — no follow-up questions.

Activation Triggers

This skill activates when:

  • flow-detector has emitted two or more candidate flows.
  • The caller (/clarify command or @clarifier agent) hands off the candidate list.

This skill does not activate when:

  • flow-detector returned exactly one candidate. The caller proceeds straight to execution-context-builder. If this skill is invoked with a single-candidate list, no-op and log "single candidate; selector skipped".

Hard Rules

  1. Issue exactly one AskUserQuestion call. Never two. Never a follow-up.
  2. The question's options are the candidate titles. The "Other" escape hatch is added automatically by AskUserQuestion.
  3. Do not ask sub-questions about entity, trigger, channel, recipient, or failure handling. Those are encoded in each candidate's full_context.
  4. Do not ask "are you sure?" after the user picks.
  5. Pass the chosen candidate's full_context directly to execution-context-builder.

Process

1. Print the framing

Output a brief framing block before the question:

This task has multiple possible flows:

1. <candidate[0].title>
   <candidate[0].one_line_diff>

2. <candidate[1].title>
   <candidate[1].one_line_diff>

…

Which flow would you like to proceed with?

The framing prints to the chat so the user can see all candidates at once. The AskUserQuestion tool then renders them as selectable options.

2. Issue the question

Call AskUserQuestion with one question:

question: "Which flow describes the intended scope?"
header: "Flow"
multiSelect: false
options:
  - label: <candidate[0].title (truncate to ≤ 5 words if needed)>
    description: <candidate[0].one_line_diff>
  - label: <candidate[1].title>
    description: <candidate[1].one_line_diff>
  # … one entry per candidate, max 4 total

Notes:

  • AskUserQuestion allows 2–4 options. flow-detector caps at 4 candidates, so this fits without truncation.
  • The "Other" escape hatch is added automatically. Do not include an explicit "Other" option.

Read the full file on GitHub · 125 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. 4d ago First seen · 125 lines · 49 tokens per session scan A e7a3124a5f5e

Subscribe to this mod's changes

flow-selector is a skill published in the GitHub repository smicolon/ai-kit (6 stars, last pushed yesterday), licensed MIT. It adds 49 tokens to every session and 1,209 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.