Line 7
Official@line7works · San Francisco
Line 7 provides a Claude Code skill suite for slice-based software builds, adversarial code review, project lifecycle management, and multi-model consensus vetting.
Agent Skills by Line 7
Showing 24 vetted skills indexed across 1 GitHub repositories.
line7workssunrise
Bootstrap a new project across repo, GitHub, Vercel, database, Notion, and vault layers.
line7workssunset
Archive abandoned projects across vault, repo, GitHub, Vercel, database, and Notion layers.
line7worksdelivery-probe
Echoes sentinel lines from its own body to verify skill delivery.
line7worksmanual-only-probe
Replies with the fixed line PROBE-RAN when explicitly invoked by the user.
line7worksrecheck-v2
Verifies recorded build fixes with a fresh closed-checklist verifier and moves slice status cards.
line7worksinspect
Reviews build plan documents adversarially through fresh inspector models before construction begins.
line7workswargame
Generates adversarial pre-mortem plans with ranked failure modes, kill criteria, and verified checks.
line7worksshutdown
Writes verified session handoff files and memory pointers before terminal restarts.
line7worksarcade
Publish, update, reorder, and delete pages on the Line 7 Arcade via the bundled arcade-publish CLI.
line7workshandoff
Writes dated end-of-slice handoff blocks and kickoff lines into build docs.
line7workssignoff
Runs independent adversarial review of freshly built code slices and issues a signed verdict.
line7worksdigest
Compile a repo's raw feedback log into a current-state notes.md board.
line7worksfb
Appends verbatim feedback notes to the correct docs/feedback.md inbox file.
line7workshuh
Restates pending questions or findings in plain language with options and a recommendation.
line7worksprecon
Harvests idea discussions into fixed-format scope docs for downstream build planning.
line7worksarchitect
Produces an architecture-and-delivery doc from a precon scope doc through a structured interview.
line7worksforge
Generate, inspect, and refine AI images on Fal.ai through a Python CLI with cost caps.
line7worksprint-tune
Guides collaborative per-setting tuning of 3D print jobs for the Bambu H2C printer.
line7worksreaders
Dispatches cold, read-only document reviews to fresh Claude, GPT, Gemini, DeepSeek, or Qwen readers.
line7worksship
Runs one build slice through build, signoff, fix, and recheck stations with a hard lap limit.
line7worksjpb
Vets product ideas by collecting independent product boxes from multiple frontier AI models.
line7worksvertical
Runs whole-build adversarial review with local and outside model reviewers, producing one verified verdict doc.
line7worksblueprint
Drafts a build document in dependency-ordered, verifiable slices from a feature discussion.
line7worksrecheck
Verifies named signoff fixes landed and flips the slice status card.
Frequently Asked Questions About Line 7
FAQPage SchemaWhat tasks can I accomplish with Line 7's skills?▼
You can bootstrap and archive full projects (sunrise/sunset), draft sliced build docs (blueprint), execute slices (build), run adversarial reviews (signoff, inspect, vertical, wargame), verify fixes (recheck), capture feedback (fb, digest), manage session handoffs, publish pages to the Line 7 Arcade, generate images via Fal.ai, and tune Bambu H2C 3D prints.
Who are Line 7's skills designed for?▼
They are built for a solo developer (Tony) running Claude Code in a terminal, managing multiple repos under ~/Developer. The skills suit engineers who want disciplined, slice-by-slice builds with independent review gates, documented handoffs between sessions, and reversible project lifecycle operations.
How do the build-loop skills work together in practice?▼
A typical run is /precon to harvest scope, /architect for the delivery doc, /blueprint to slice the plan, then /build per slice, /signoff for adversarial review, and /recheck to verify fixes. The /ship command chains build, signoff, fix, and recheck for one slice automatically; /vertical signs off the whole build.
What external services and dependencies do the skills require?▼
Depending on the skill: GitHub, Vercel (with Marketplace), Supabase or Neon databases, Notion, and Obsidian for project lifecycle; Fal.ai for image generation; a Bambu H2C printer for print-tune; and access to outside models (GPT, Gemini, DeepSeek, Qwen) for readers and jpb consensus runs.
Are destructive actions safe with Line 7's skills?▼
Yes. The sunset skill never deletes anything — every archival step is reversible, previews all changes first, and requires a one-line reason. The arcade skill writes to live production but demands an explicit conversational yes before any delete. Shutdown is read-only against git and never commits.