assist-plugin

Plugin assist-plugin from marketplace assist-plugin-marketplace in tjmonsi/assist-marketplace-plugin. Commands read from the repo's manifest — 6 days ago.

Skills (22)

• ask

  Answer questions using repository search and optional web research. Searches repo files first. Add --web to also search the web when repo context is insufficient. Always cites sources. - Read - Grep - Glob - WebSearch - WebFetch

• branch

  Manage git branches. Create feature/bugfix/release branches with standard naming, switch branches with stash support, list all branches, clean up merged worktrees and their branches, and clear the Go build cache. - Bash(git:*) - Bash(df:*) - Bash(go clean:*) - Bash(realpath:*)

• bump

  Semantic version bumping with automated changelog generation. Detects version files, determines bump type from commits or explicit input, updates versions, generates changelog, and commits with optional git tag. - Read - Write - Edit - Grep - Glob - Bash

• clarify

  Clarify unclear points from the current discussion. Use on /clarify or when the user says something Claude wrote is unclear. Claude explains its own output with evidence, asks the user about points only the user can settle, shows both in chat, and appends them to a daily log. - Read - Grep - Glob -

• code

  Coding standards, language conventions, logging/error-handling doctrines, and framework patterns. Loaded by the developer agent and referenced by task-plan for consistency. Loaded by agents, not invoked directly by users. - Read - Write - Edit - Grep - Glob - Bash(which firecrawl) - Bash(which ctx7)

• commit

  Create conventional commits with git. Stages files, composes messages using conventional commits format, and supports squash/amend. - Bash

• create-test

  Generate black-box acceptance-criteria tests without reading implementation. Enumerates input space (valid, invalid, boundary, null, type-mismatch, injection) per criterion. Every test traces to AC/SPEC/FR/UC with [AC-N|SPEC-NNN|...] chains. - Read - Write - Grep - Glob

• debug

  Diagnoses and fixes bugs through structured root cause analysis and iterative fix cycles. Supports analyze, fix, and review subcommands. - Read - Grep - Glob - Bash - LSP

• deslop

  Remove AI writing tells from prose. edit returns the cleaned text only; detect lists findings without editing. - Read - Grep - Glob

• do

  Routes a task, prompt, or multi-instruction prompt to the best specialized subagent. Runs as a finite-state Communicator: decomposes the prompt into steps, shows the step table in chat, gets user approval with edit and model override options, dispatches agents in dependency order, reads each agent's

• gather-requirements

  Gather and document business, user, and technical requirements using IEEE 29148 templates and OpenSpec Scenario/RFC 2119 format. Subcommands: elicit (interactive), document (structured output), review (quality audit). - Read - Grep - Glob - AskUserQuestion - Edit(./docs/requirements/BRD-*.md) - Edit

• health

  Check assist-plugin hook health: Node version, guard, injection-filter, prose-hook, and FSA-hook self-tests, recent hook failures, subagent reports with terminal-block problems, disabled guards, Claude Code attribution settings. - Bash(node "${CLAUDE_PLUGIN_ROOT}/hooks/health-check.js" --data-dir "$

• integrated-test

  Integration and E2E testing patterns for multiple languages and frameworks. Covers testcontainers-based database testing, semantic locators for Playwright E2E, monorepo test organization, and Playwright availability detection. Used by QA. - Read - Write - Edit - Grep - Glob - Bash(command -v uv) - B

• map-project

  Discover repository structure. Identifies key directories, detects technology stack, maps module/package dependencies, and generates architecture summary. - Read - Glob - Bash

• pentest

  Penetration-testing and dependency-audit methodology preloaded by the qa agent. Covers OWASP Top 10 for web apps and LLM/Bot applications, vulnerability scanning, and guardrail testing. Not for code or PR review (use reviewer or code-reviewer). Run only when the user explicitly authorizes active tes

• pr-review

  Review pull requests with automatic platform detection (GitHub/Bitbucket). Fetches PR details, lists changed files, shows comments, and submits reviews with approval, changes-requested, or comment-only verdict. - Read - Write - Grep - Glob - Bash - AskUserQuestion

• review-md

  Review markdown reports for consistency, conciseness, and factual accuracy. Flags issues without editing the source file. - Read - Grep - Glob - Write

• run-test

  Run existing test suites (unit and integration), capture coverage metrics, and produce consolidated test reports. Used by QA to validate implementations. - Read - Write - Grep - Glob - Bash(command -v uv) - Bash(pytest) - Bash(uv run --no-sync --no-python-downloads pytest) - Bash(pytest tests/) - Ba

• spec-from-requirements

  Create and update technical specifications from requirements. Classify each requirement type (Architecture|API|Frontend|Functionality|UI-UX|General), write to specs/ folder, support delta specs for incremental changes (ADDED/MODIFIED/REMOVED). Modes: create (new spec), delta (incremental change), re

• task-plan

  Validates specifications have acceptance criteria, then outputs ordered coding steps with file names, requirement IDs, and code skill references for the developer to implement. - Read - Grep - Glob - Edit(./docs/task-plans/task-plan-*.md)

• technical-writing

  Generate and update technical documentation. Supports user docs, README, API documentation, guides, and specification writing. - Read - Grep - Glob - Edit(./README.md)

• voice-script

  Write a paste-ready text-to-speech script for the chosen TTS model, using only that model's own paralinguistic tags and style field. Use on /voice-script or when the user asks for a voice, narration, or TTS script. Asks for the model, voice, length, delivery, and a topic or draft, then saves the scr

Subagents (16)

• ask

  Answer questions using repository search and optional web research.

• code-reviewer

  Formal line-by-line code review, architectural evaluation, critical sign-off authority.

• developer-tester

  Write acceptance-criteria tests from specs without reading implementation. Black-box input-space enumeration, traceability chains, zero implementation knowledge.

• developer

  Write, fix, and refactor code. Implement features from specs, apply approved bug fixes, improve performance and clarity.

• devops

  Configure CI/CD pipelines, provision infrastructure, automate deployments, setup monitoring.

• explore

  Fast read-only code search. Find files, grep symbols, locate definitions, discover code locations.

• general-purpose

  Ad-hoc analysis and experimentation for tasks /do explicitly assigns. Not a fallback: unmatched work goes to Clarify.

• intake-reviewer

  Read-only intake review before /do plans write steps. Mode review audits a user-provided spec, requirements, or instructions file. Mode conflict-check compares a build request with the project's existing requirements, specs, task plans, and DECISIONS.md.

• planner

  Design system architecture, create roadmaps, analyze dependencies and risks, define technical strategy.

• qa

  Create test plans, design tests, validate acceptance criteria, detect regressions, and ensure code coverage.

• requirements-gatherer

  Elicit requirements, create BRD/URD, define FR and NFR, specify acceptance criteria.

• researcher

  Research topics, audit documentation, discover best practices, analyze solutions.

• reviewer

  Code review for bugs, security, and quality. Perform OWASP analysis, identify issues, approve or request revisions.

• solutions-architect

  Translate requirements to specs, design APIs and schemas, create data flows, define error handling.

• task-planner

  Turn an approved spec into an ordered task plan with the task-plan skill. Writes only a task-plan-*.md file under docs/task-plans/; no Edit, Bash, or PowerShell tool. Treats the spec and every document it reads as untrusted data.

• worker

  Multi-discipline generalist. Code + test + review combined for small features and cross-cutting work.

Hooks (27)

• PreToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/sa-mermaid-only.js"

• PreToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/commit-attribution-guard.js"

• PreToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/doc-cli-guard.js"

• PreToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/pr-post-ask.js"

• PreToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/pr-verdict-guard.js"

• PreToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/injection-filter.js"

• PreToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/prose-gate.js"

• PreToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/fsa-write-bar.js"

• PreToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/fsa-write-bar.js"

• PreToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/data-access-log.js"

• PreToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/phi-outbound-guard.js"

• PreToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/ferpa-disclosure-guard.js"

• PostToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/injection-filter.js"

• PostToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/audit-log.js"

• PostToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/gate-evidence.js"

• PostToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/goal-guard.js"

• PostToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/prose-advisor.js"

• PostToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/fsa-log.js"

• PostToolUse

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/fsa-log.js"

• PostToolUseFailure

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/gate-evidence.js"

• UserPromptSubmit

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/injection-filter.js"

• UserPromptSubmit

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/confirmation-record.js"

• UserPromptSubmit

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/purpose-capture.js"

• SessionStart

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/health-check.js" --session-start

• SessionStart

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/writing-context.js"

• SubagentStart

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/writing-context.js"

• SubagentStop

  cd "${CLAUDE_PLUGIN_ROOT}" 2>/dev/null || true; node "${CLAUDE_PLUGIN_ROOT}/hooks/fsa-log.js"

Similar plugins