auto
Run an increment unattended - the Stop hook feeds you back into the loop until every task is done. Use when saying "auto mode", "run until done", or "ship this…
Open a SpecWeave increment - a numbered folder with spec.md (problem, scope, numbered ACs, approach) and tasks.md - before any code. Use for "plan a feature", "new increment", "let's build X".
$ npx -y skills add anton-abyzov/specweave --skill sw-increment --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/sw-incrementContext preview
The summary Claude sees to decide when to auto-load this skill.
Open a SpecWeave increment - a numbered folder with spec.md (problem, scope, numbered ACs, approach) and tasks.md - before any code. Use for "plan a feature", "new increment", "let's build X".
description: Open a SpecWeave increment - a numbered folder with spec.md (problem, scope, numbered ACs, approach) and tasks.md - before any code. Use for "plan a feature", "new increment", "let's build X". argument-hint: "\"<feature title>\" [--supersedes NNNN]" version: "2.0.0"
An increment is one folder, `.specweave/increments/NNNN-slug/`, and it is the tracker: spec + tasks + ledger + evidence live together and travel in git. Write the spec, get it approved, then implement (`sw-do`). No code before the ACs exist.
Run `specweave --version`: exit 0 → **CLI path**. Not found → **manual path**.
Ask only what you cannot infer from the repo — typically: the problem behind the request, what is explicitly out of scope, and how you will know it works. Scope clear from the request? Skip the interview and write the spec.
CLI:
specweave create-increment --auto-id --name "<slug>" --title "<Title>" \ --description "<one line>" --project "<project id from .specweave/config.json>" \ [--type feature|bug|hotfix|refactor|experiment]
Manual: next id = highest existing number + 1, zero-padded to 4 (`ls .specweave/increments`), folder `NNNN-kebab-slug`, then write `metadata.json`, `spec.md`, `tasks.md`, and an empty `ledger.jsonl`:
{"id":"0042-ledger-fold","status":"active","type":"feature","created":"2026-09-02T10:00:00Z","lastActivity":"2026-09-02T10:00:00Z"}Write every file as UTF-8 **without a BOM** (PowerShell: `[IO.File]::WriteAllText($p, $body, [Text.UTF8Encoding]::new($false))`, never `>`) — a BOM breaks the parsers that read these files on the next machine.
Status is one of `planning` `active` `paused` `ready_for_review` `completed` `abandoned`, and only lifecycle commands change it. Replacing an older increment? Add `"supersedes":"0031-old-slug"` here and close that one as abandoned with the reason `superseded by 0042-ledger-fold`.
# 0042 Ledger fold ## Problem What is wrong today and for whom. The intent, not the solution. ## Scope In: … · Out: … (say what you are deliberately not doing) ## Acceptance criteria - [ ] AC-01 Two agents claiming the same task never both get it - [ ] AC-02 A malformed ledger line is skipped and counted, never fatal ## Approach Files that change and in what order · risks · decisions taken (link ADRs) · alternatives rejected and why. ## Open questions - …
ACs are numbered, checkable, and each one is testable by a command or an observation. `plan.md` is optional overflow when the Approach outgrows the spec.
# Tasks — 0042 Ledger fold ### T-01 Fold the ledger - AC: AC-01 | Files: src/core/tasks/ledger.ts, src/core/tasks/ledger.test.ts | Test: npm test -- ledger ### T-02 Skip malformed lines - AC: AC-02 | Files: src/core/tasks/ledger.ts | Test: npm test -- ledger
Show the ACs and the task list, get an explicit go-ahead, then run `sw-do`. Set `metadata.json.status` to `active` when work starts.
`specweave complete` refuses to close without passing evidence.
Spec-first AI development: describe a feature → AI creates spec + plan + tasks, builds autonomously, syncs to GitHub/JIRA. Domain-expert skills for PM, Architect, Frontend, QA learn your patterns permanently. Claude Code, Codex, Cursor, Copilot & more.
Repo: anton-abyzov/specweave
Run an increment unattended - the Stop hook feeds you back into the loop until every task is done. Use when saying "auto mode", "run until done", or "ship this…
Expand the solution space before committing - framed options compared on stated criteria, ending in a pick. Use when saying "brainstorm", "ideate", or "what…
Work an increment task by task through the ledger: task next, claim, implement, commit, task done with evidence. Use when saying "implement", "start working",…
Close an increment: ledger check, specweave verify, optional review, then specweave complete. Use when all tasks are done and saying "close increment", "we are…
Write a portable, secret-scrubbed handoff doc so this work can continue in any AI tool or on any machine. Use when saying "handoff", "running out of tokens",…
Plan a unit of work as a SpecWeave increment - spec.md with Problem, Scope, numbered ACs and an Approach, plus tasks.md. Use when starting a feature, bug,…