agentleFS
anivaranivar/agentic-sdlc/skills/cut-tickets/SKILL.md

cut-tickets

Skill0 starsChanged 37 days agoMITView on GitHub

Break an approved plan, spec or conversation into tickets: one per real work item, each a vertical slice with acceptance criteria and the tickets that block it, free of infrastructure identifiers, secret paths, partner URLs and colleagues' names, with full links for every reference. The breakdown is shown and approved in the conversation; publishing to the tracker is a post and waits for its go-ahead. Triggers on: "cut tickets", "break this down", "create issues", "to tickets", "split into tasks", "what are the work items", "file the issues", "ticket this".

The file

Files
SKILL.md

Cut tickets

A ticket is a promise someone else can check. Cut them so each one is a complete, narrow path through every layer, demoable or verifiable on its own, and so the set as a whole has a frontier: the tickets whose blockers are done.

The rules

IdRuleIn practice
TICKET-1One ticket per real work itemNever one per checklist step. A vertical slice through schema, API, interface and tests, sized to one working session (one fresh context window for an agent). Any prefactoring goes first.
TICKET-2Acceptance and blockersWhat to build from the user's perspective, acceptance criteria someone else can check, the tickets that block it. Native dependency links where the tracker has them; a "Blocked by" line otherwise.
TICKET-3No identifiers, secrets or namesNo infrastructure identifiers, secret paths, partner URLs or colleagues' names. Describe the step and point to the internal runbook. The owner is a team or role.
TICKET-4Full linksEvery reference to another change or ticket is a full link, never a bare number.
TICKET-5Approval before publishingShow the breakdown as a numbered list: title, blocked by, what it delivers. Ask about granularity, blocking edges, merges and splits. Publish only after approval, and publishing is a post (ASK-3); on publish, apply the tracker's ready-for-agent label unless told otherwise (docs/agents/tracker.md). The parent item is never closed or edited by the breakdown.
TICKET-6Wide refactors use expand and contractA mechanical change that breaks thousands of call sites at once cannot be a slice. Expand (add the new form beside the old), migrate in batches sized by blast radius, contract (delete the old form). Each batch its own ticket, blocked by the expand; the contract blocked by every batch.

Where tickets go

The tracker and its label vocabulary are recorded in docs/agents/tracker.md (SETUP-2). If the file is absent, say so and ask the decider to run /setup-conventions; until then write local tickets under .scratch/<feature>/issues/NN-slug.md and add .scratch/ to .gitignore so a local ticket is never one git add away from the public repository.

Ticket template

See references/ticket-template.md. Avoid file paths and code in a ticket: they go stale. The exception is a snippet from a prototype that encodes a decision more precisely than prose (a state machine, a schema, a type shape); inline the decision-rich part and say where it came from.

Companion skills

NeedSkill
The plan the tickets come fromplan-project
The voice of the ticket textplain-voice
Landing each ticketthin-slices
Recording the tracker and labelssetup-conventions

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.