claude

proof

Spec -> build -> review -> ship, with a running record of decisions. Configured per project.

10 on GitHubby Proofv0.11.0MITupdated 29 days ago
/

Plugin proof from marketplace proof in jsamitt/proof-claude-plugin. Commands read from the repo's manifest — 29 days ago.

This plugin runs code on your machine — its hooks fire automatically on events like tool use. The exact commands are listed below; read them before installing.

What it installs

Skills (11)

  • proof-design

    Use when setting up or checking a project's design system — not part of every issue. Establishes a design system where a project has none, wire an existing one into the harness so the other skills obey it, or report drift between the tokens and the code — detects which of the three applies rather th

  • proof-fix

    Use when something is broken — finds the root cause before changing anything, confirms it, fixes it with a test that proves it, and opens a PR. The path every bug takes, instead of spec and work. Part of every bug

  • proof-handoff

    Use when a design made in a design tool is ready to be built — checks it against the codebase before anything is built (what exists, what is new, what cannot be built as drawn, which states are missing), writes the revisions for the design tool, repeats until it is clean, then records the approved d

  • proof-init

    Use once per project — not part of every issue — when setting up the Proof harness for the first time, or when its configuration is out of date — inspects the repo, asks what it cannot infer, and writes .claude/harness.json plus the GitHub board and labels

  • proof-learn

    The last step of every issue, and whenever a working session pauses — sweeps the session for lessons about the product and code while the conversation still exists. Owns what the project knows: records a lesson directly, searches past decisions, promotes durable ones into the decisions file behind a

  • proof-plan

    Use periodically — not part of every issue — to decide what to work on next and in what order, re-check priorities that have gone stale, and optionally clean up duplicate, superseded or abandoned issues. Proposes; changes nothing without approval

  • proof-retro

    Use periodically — monthly or after a release, not part of every issue, or let the retro routine run it. Owns how the work goes, not what was learned. Looks back over a stretch of work: shipped against planned, where work waited, how estimates held up, which areas keep breaking, how often lessons re

  • proof-review

    Use when a pull request or working diff needs review — runs three independent reviewers (correctness, adversarial breaker, project risk) in parallel and posts consolidated findings. Part of every issue

  • proof-ship

    Use when merged work needs to go out — prepares the release (version, changelog, app-store notes) as its own pull request, then runs the project's preflight and deploy steps in order, halting on the first failure, and closes out the board. Part of every release

  • proof-spec

    Use when turning an idea into something buildable — first checks it against the board for anything it overlaps, supersedes or contradicts and settles each, then explores the problem, proposes approaches, writes user stories and acceptance criteria, rates priority, asks whether a new idea lands as a

  • proof-work

    Use when building a specced GitHub issue end to end — plans, pauses once for approval, then implements, verifies, records learnings, opens a PR and updates the board without further interruption. Part of every issue

Subagents (5)

  • proof-builder

    Builds an approved plan for /proof-work or /proof-fix on the model the developer chose — implements and commits criterion by criterion, and hands back a report or a question rather than deciding anything the plan did not cover

  • proof-review-breaker

    Adversarial reviewer — assumes the change is broken and tries to prove it with concrete failing inputs, boundary values, and sequences the happy path never exercises

  • proof-review-correctness

    Reviews a diff for correctness, structure and maintainability — does it do what the issue asked, does it hold up under the codebase's own conventions, will it be understandable in six months

  • proof-review-design

    Reviews a diff for design-system adherence and platform accessibility — cites the project's own tokens and rules rather than general taste, and says so plainly when there is nothing to cite

  • proof-review-risk

    Reviews a diff against the project's own declared risk areas — the things that would be genuinely bad if they broke, as named in the project config rather than assumed from a generic checklist

Hooks (7)

  • SessionStart

    "${CLAUDE_PLUGIN_ROOT}/scripts/hooks/update-check.sh"

  • PreToolUse

    "${CLAUDE_PLUGIN_ROOT}/scripts/hooks/pre-bash-guard.sh"

  • PreToolUse

    "${CLAUDE_PLUGIN_ROOT}/scripts/hooks/pre-write-guard.sh"

  • PostToolUse

    "${CLAUDE_PLUGIN_ROOT}/scripts/hooks/session-branch.sh" record

  • PostToolUseFailure

    "${CLAUDE_PLUGIN_ROOT}/scripts/hooks/session-branch.sh" record

  • Stop

    "${CLAUDE_PLUGIN_ROOT}/scripts/hooks/stop-verify.sh"

  • PreCompact

    "${CLAUDE_PLUGIN_ROOT}/scripts/hooks/precompact-archive.sh"

View source on GitHub

Similar plugins

superpowers

claudecodex

Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques

15 skills · hooks

obra/superpowers 316,836

diagram-design

claudecodex

Editorial diagrams in your brand, as self-contained HTML and SVG. Architecture, flowcharts, sequence, state machines, ER and database schemas, timelines, swimlanes, org charts, Gantt, Wardley maps, user journeys, service blueprints, UML class, lifecycle phase maps, and charts from bar, line and scatter to Sankey, treemap and heatmap, plus many more. Redraws draw.io, Mermaid, Excalidraw and PlantUML files. Exports SVG and PNG.

1 skill · 7 commands

cathrynlavery/diagram-design 66,757

Backend API design, GraphQL architecture, workflow orchestration with Temporal, and test-driven backend development

9 skills · 1 command · 8 agents

wshobson/agents 42,461

beads

claudecodex

AI-supervised issue tracker for coding workflows

1 skill · 1 agent · hooks

gastownhall/beads 28,698

MiniMax AI skills library for frontend, fullstack, Android, iOS, shader, GIF sticker, document, presentation, spreadsheet, and multimodal media workflows

18 skills

MiniMax-AI/skills 13,828

worktrunk

claudecodex

Worktrunk is a CLI for Git worktree management, designed for parallel AI agent workflows. This plugin provides configuration guidance (LLM commit messages, project hooks, worktree paths), automatic activity tracking (🤖/💬 indicators in `wt list` showing active Claude sessions), and a `/wt-diff` pane showing each worktree's changes.

2 skills · hooks

max-sixty/worktrunk 12,481

Maintain proof?

This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.

claude & codex pluginslisted