sprint-loop
/sprint-loop
Run or resume the Sprint Loops Book v2 workflow when the user explicitly invokes /sprint-loop or directly asks to start, continue, resume, or run a sprint loop. Do not use it for ordinary documentation work or merely because a docs directory exists.
- Author
- crussella0129 ↗
- Repository
- animus_sprint_loops ↗
- Licence
- MIT
- Last repository change
- 23.09.2026
Assessment and security
- Files in this package
- 55
- Scripts
- 35
- HTML comments
- 0
- URLs in SKILL.md
- 0
No signals detected. Repository changed on 23.09.2026, without references to retired models or APIs.
MIT, open: use and modify while retaining author attribution.
- 35 scripts in this package. Review them before installing, as Claude may execute them.
Scripts: scripts/abort-sprint.sh, scripts/book-paths.sh, scripts/book-routing.test.sh, scripts/bundle-version.sh, scripts/check-book.sh, scripts/check-book.test.sh, scripts/check-substrate.sh, scripts/check-substrate.test.sh, scripts/check-tracked.sh, scripts/check-tracked.test.sh, scripts/close-sprint.sh, scripts/commit-task.sh, scripts/critic-contract.sh, scripts/current-phase.sh, scripts/current-sprint.
What it does and when to use it
Run or resume the Sprint Loops Book v2 workflow when the user explicitly invokes /sprint-loop or directly asks to start, continue, resume, or run a sprint loop. Do not use it for ordinary documentation work or merely because a docs directory exists.
Arguments
/sprint-loop [ continue | start <goal> | abort <reason> | upgrade ]
Original author description
Run or resume the Sprint Loops Book v2 workflow when the user explicitly invokes /sprint-loop or directly asks to start, continue, resume, or run a sprint loop. Do not use it for ordinary documentation work or merely because a docs directory exists.
How it works
- Activation
- Resolve roots
- Arguments
- Route
- Book and Claude Code authority
SKILL.md preview
---
name: sprint-loop
description: Run or resume the Sprint Loops Book v2 workflow when the user explicitly invokes /sprint-loop or directly asks to start, continue, resume, or run a sprint loop. Do not use it for ordinary documentation work or merely because a docs directory exists.
argument-hint: "[ continue | start <goal> | abort <reason> | upgrade ]"
---
# Sprint Loops for Claude Code
This adapter activates and routes the shared Sprint Loops workflow. The
selected phase file owns its phase contract; the Project Book owns durable
project meaning and state.
## Activation
Honor the frontmatter boundary. A Book marker may confirm a direct resume
request, but filesystem or documentation presence is never an activation
signal by itself.
## Resolve roots
- `${CLAUDE_SKILL_DIR}` is the absolute directory of this loaded skill.
Require it, and quote every bundled path resolved beneath it.
- `<project-root>` is the target repository's Git top level or, outside Git,
the active workspace root. Never treat the skill directory as the project.
Resolve both once. If the target project is ambiguous, ask before writing.
Run every helper with `<project-root>` as the working directory.
## Arguments
The invocation arguments are `$ARGUMENTS`.
- No argument or `continue` resumes the state reported by the router.
- `start <goal>` retains `<goal>` as the requested sprint objective. It may
initialize only when the router reports `uninitialized` or
`ready-for-next-sprint`; resume or explicitly abort an active sprint first.
- `abort <reason>` invokes
`bash "${CLAUDE_SKILL_DIR}/scripts/abort-sprint.sh" "<reason>"`. Require one
non-empty, one-line reason.
- `upgrade` converges the project's substrate to this bundle's contract
version without starting a sprint: run
`bash "${CLAUDE_SKILL_DIR}/scripts/check-substrate.sh"`, then
`bash "${CLAUDE_SKILL_DIR}/scripts/deploy-substrate.sh"`, then report the
resulting substrate state. Spin-up and upgrade are the same idempotent
command, so this is safe on a current project — it changes nothing. Add
`--check` to the second command to name pending steps without writing.
- Any other argument is unsupported. Never use an argument to jump directly
to a phase.
## Route
From `<project-root>`, run:
```bash
bash "${CLAUDE_SKILL_DIR}/scripts/current-phase.sh"
```
Load only the matching contract:
- `uninitialized` or `ready-for-next-sprint` →
`${CLAUDE_SKILL_DIR}/phases/01-init-sprint.md`
- `research` → `${CLAUDE_SKILL_DIR}/phases/02-research-phase.md`
- `plan` → `${CLAUDE_SKILL_DIR}/phases/03-plan-phase.md`
- `build` → `${CLAUDE_SKILL_DIR}/phases/04-build-phase.md`
- `test` → `${CLAUDE_SKILL_DIR}/phases/05-test-phase.md`
- `loop` → `${CLAUDE_SKILL_DIR}/phases/06-loop-phase.md`
Read `${CLAUDE_SKILL_DIR}/phases/00-overview.md` only to interpret the Book
contract or a layout diagnostic. Resolve legacy-only state through
`${CLAUDE_SKILL_DIR}/scripts/migrate-to-book.sh`; never create a second
writable state tree. After satisfying Exit evidence, re-run the router and
stop at the boundary the user requested.
## Book and Claude Code authority
Book schema v2 keeps semantic intent in `docs/intents/`, work state in
`docs/work/`, sprint provenance in `docs/sprints/`, and navigation-only
views in `docs/SUMMARY.md`. The filesystem is the state machine; lower-
authority evidence cannot silently redefine intent.
`${CLAUDE_SKILL_DIR}/phases/03-plan-phase.md` owns Claude Code Plan Mode.
For session-scoped recurrence, the user may start
`/loop /sprint-loop continue`; each invocation routes from Book evidence,
and the user starts and stops recurrence. Plan Mode, plan approval,
auto-accept, and `/loop` affect orchestration only: none enlarges authority,
changes active permissions, bypasses a phase gate, or validates unverifiable
evidence. Do not weaken permission or security controls to keep a loop moving.
Push, merge, release, force-push, delete, and material scope expansion require an explicit request or a declared preauthorized-remote profile. The remote profile is declared per `schemas/remote-profile.md` and drives the one-PR/MR-per-sprint `work -> base` checkpoint (human-approve by default); the substrate gate `scripts/check-substrate.sh` establishes and verifies it, and the skill creates no per-sprint branch.
Files in this package (55)
- SKILL.md 4 KB
- phases/00-overview.md 3 KB
- phases/01-init-sprint.md 6 KB
- phases/02-research-phase.md 2 KB
- phases/03-plan-phase.md 3 KB
- phases/04-build-phase.md 3 KB
- phases/05-test-phase.md 3 KB
- phases/06-loop-phase.md 6 KB
- prompts/plan-critic.md 3 KB
- prompts/test-critic.md 3 KB
- schemas/agent-tasks.md 1 KB
- schemas/build-plan.md 2 KB
- schemas/completed-tasks.md 1 KB
- schemas/failure-report.md 1 KB
- schemas/intent.md 3 KB
- schemas/remote-profile.md 3 KB
- schemas/research-report.md 2 KB
- schemas/sprint-meta.md 2 KB
- schemas/test-plan.md 1 KB
- schemas/test-report.md 2 KB
- scripts/abort-sprint.sh 3 KB
- scripts/book-paths.sh 8 KB
- scripts/book-routing.test.sh 8 KB
- scripts/bundle-version.sh 1 KB
- scripts/check-book.sh 4 KB
- scripts/check-book.test.sh 4 KB
- scripts/check-substrate.sh 4 KB
- scripts/check-substrate.test.sh 8 KB
- scripts/check-tracked.sh 2 KB
- scripts/check-tracked.test.sh 4 KB
- scripts/close-sprint.sh 6 KB
- scripts/commit-task.sh 3 KB
- scripts/critic-contract.sh 2 KB
- scripts/current-phase.sh 1 KB
- scripts/current-sprint.sh 1 KB
- scripts/deploy-substrate.sh 17 KB
- scripts/deploy-substrate.test.sh 31 KB
- scripts/detect-languages.sh 2 KB
- scripts/detect-languages.test.sh 3 KB
- scripts/finalize-plan.sh 7 KB
- scripts/init-sprint.sh 5 KB
- scripts/migrate-to-book.sh 28 KB
- scripts/migrate-to-book.test.sh 17 KB
- scripts/remote-adapter.sh 7 KB
- scripts/remote-adapter.test.sh 13 KB
- scripts/remote-profile.sh 3 KB
- scripts/remote-profile.test.sh 6 KB
- scripts/research-budget.sh 1 KB
- scripts/runtime-helpers.test.sh 29 KB
- scripts/scaffold-ci.sh 8 KB
- scripts/scaffold-ci.test.sh 8 KB
- scripts/selftest.sh 1 KB
- scripts/sync-work-branch.sh 2 KB
- scripts/sync-work-branch.test.sh 2 KB
- scripts/update-confidence.sh 1 KB
Versions
The database has no other skills named sprint-loop.
Database snapshot: 13.11.2026 · package version e12e90a963fa (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 100sprint-loops
Run or resume the Sprint Loops Book v2 workflow when the user explicitly invokes $sprint-loops or directly asks to start, continue, resume, or run a sprint loop. Do not use it for ordinary documentation work or merely because a docs directory exists. - 100sprint
Sprint workflow pipeline — chains plan → build → test → review → ship skills into a structured sprint. Use when starting a new feature or project iteration to follow the full lifecycle. - 100sprint-runner
Use when the user wants to RUN a sprint as a facilitated workshop with a virtual team — not just get sprint-style analysis. Triggers: 'run a foundation sprint', 'run a design sprint', 'sprint runner', 'create a virtual team for my idea', 'I want to facilitate a sprint with… - 100sprint-evaluator
Evaluate sprint implementations against design docs and acceptance criteria. Runs after sprint-executor to assess quality with concrete scoring rubric (100 points, 70 to pass). Provides actionable feedback loop on failure. Use when user says "evaluate sprint", "review… - 100sprint-retro
Use only when the user explicitly asks to run a sprint retrospective or team reflection on what went well and what to change. - 100gds-sprint-planning
Generate sprint status tracking from epics. Use when the user says "run sprint planning" or "generate sprint plan