start

A guided starting command for software work. It checks whether related work already exists and routes you based on whether you have a Jira ticket, want to investigate directly, or need to create a task first.

In plain words
What is it for?
Use it as the first step for a new request, investigation, or Jira-linked change to select the appropriate project path.
Why use it?
It prevents new work from starting in the wrong workflow or duplicating an existing task or change.

Command 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 commands/iolairus/borderlint/start
Clone the repo
git clone --depth 1 https://github.com/iolairus/borderlint

Made for: Claude Code.

Per session 18 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 623 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 $0.00018 $0.00623
Opus 5 $0.00009 $0.00311
Sonnet 5 $0.00004 $0.00125
Haiku 4.5 $0.00002 $0.00062

Measured 2d ago against content hash f197f59fde1a, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

start 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 2d 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.

.claude/commands/start.md · 31 lines

What it actually says

Single entry point for new work. Figure out which of the three starting situations applies and route the user there. $ARGUMENTS may hint at the topic or a Jira key; use it to preselect answers, never to skip the questions.

Steps

  1. Quick state check (fast, no deep inspection): git branch --show-current, in-flight changes in openspec/changes/ (excluding archive/), tasks in backlog/tasks/ with status other than done.

    • If $ARGUMENTS matches an existing task or change, or the current branch maps to an in-flight change → this is not new work. Run the /next logic instead and stop.
    • If there is other unrelated in-flight work, mention it in one line ("FYI: PROJ-123 is in progress") but continue — starting something new is allowed.
  2. Entry question. Ask with AskUserQuestion: "How does this work start?"

    • A. A Jira ticket already exists — the work is already defined in Jira.
    • B. No ticket — investigate or propose directly — technical change or exploration; no backlog tracking needed (it can be linked later).
    • C. No ticket — create the task first — the work must exist in the backlog/Jira before any spec or code.
  3. Route by answer. The user's selection counts as asking to proceed, so run the routed command's logic directly:

    • A → ask for the Jira key (e.g. PROJ-123) if not in $ARGUMENTS, then run /task-import <key>.
    • B → ask one follow-up with AskUserQuestion: does the user already know what to change?
      • Yes, the change is clear → /opsx:propose <name> (ask for a short kebab-case change name).
      • No, needs investigation first → /opsx:explore, reminding that it ends by suggesting /opsx:propose.
    • C → ask one follow-up with AskUserQuestion: initiative or single task?
      • Initiative (several tasks, requirements unclear) → /req-capture <topic>, then /task-generate.
      • Single task (one verifiable outcome, scope clear) → /task-new <title>.
  4. Never skip the gates downstream. Whatever the route, the golden rule holds: no code until an OpenSpec change exists, is reviewed, and the branch gate is resolved. /start only chooses the on-ramp.

Output format: after routing, one line stating the situation chosen and the command now running. If the user aborts the questions, list the three routes with their commands so they can invoke one manually.

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. 2d ago First seen · 31 lines · 18 tokens per session scan A f197f59fde1a

Subscribe to this mod's changes

start is a command published in the GitHub repository iolairus/borderlint (7 stars, last pushed 9d ago), licensed MIT. It adds 18 tokens to every session and 623 once invoked, about $0.0001 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.