ai-scrum

Plugin ai-scrum from marketplace shakhoff-claude-marketplace in shakhovskiya-create/shakhoff-claude-marketplace. Commands read from the repo's manifest — 1 month ago.

Skills (25)

• ac-validator

  Validates Acceptance Criteria quality (specific, testable, user-focused) and verifies AC fulfillment after task completion. Used by dor-validator (quality check) and sprint-executor (verification before task close).

• checkin-validator

  Validates honesty of /sprint:checkin claims by cross-checking against tool history. Detects fake claims (claiming step done without evidence), fake tests (assertions don't match test cases), and circular mocks. Used after each /sprint:checkin invocation.

• codex-external-reviewer

  External review via Codex CLI (cross-model second opinion). Activated if codex-plugin-cc installed. Falls back gracefully if Codex unavailable. Provides VERDICT: AGREE / DISAGREE / FOUND_NEW_ISSUES.

• context-investigator

  Investigates project context before task decomposition. Reads HANDOFF.md, project context briefs, glossary, README, recent commits, project structure to understand domain and patterns. Used by sprint-orchestrator at start of every sprint cycle.

• design-system-detector

  Detects project design system for HTML report styling. Searches by convention paths, falls back to neutral default theme. Returns design tokens (colors, typography, spacing) for html-report-generator.

• dod-validator

  Validates Definition of Done before sprint can close. Checks mandatory criteria (steps marked, goal addressed, increment exists, HANDOFF updated, test compliance, internal review) and conditional criteria. HARD GATE - blocks /sprint:complete if criteria fail.

• dor-validator

  Validates Definition of Ready for tasks before sprint execution. Checks mandatory criteria (title, description, AC, test approach, dependencies) and conditional criteria (PRD link, design link, expected ADR). Used by sprint-orchestrator before each sprint starts.

• handoff-updater

  Updates HANDOFF.md with sprint outcomes. Required as part of DoD - sprint cannot close without HANDOFF update. Validated by handoff-keeper subagent (fresh context) for meaningful content. Runs at end of each sprint and at task cycle completion.

• html-report-generator

  Generates self-contained HTML report from sprint cycle data. Picks up design system tokens from project's design system if detected (via design-system-detector). Self-contained file with inline CSS/JS, no CDN dependencies. Used at end of task cycle.

• lsp-bridge

  Universal wrapper for Language Server Protocol integration. Detects LSP MCPs (typescript-language-server-mcp, pyright-mcp, gopls-mcp, etc.). Falls back to CLI tools (mypy, tsc, go vet) if MCP unavailable. Used by dor-validator, senior-engineer-reviewer, dod-validator for type checking and references

• methodology-parser

  Parses methodology sections from skill SKILL.md files. Extracts mandatory/optional steps, success criteria, can_skip_if conditions. Used by task-decomposer to generate sprint backlog from caller skill methodology.

• plan-importer

  Imports external planning artifacts (superpowers writing-plans, prd-writer outputs, manual markdown) and extracts structured information for ai-scrum's task-decomposer. Detects format heuristically. Does NOT use external plan as final decomposition - instead extracts rich input that task-decomposer

• retrospective-writer

  Writes honest retrospectives at end of each sprint and at task cycle completion. Used by retrospective-author subagent (fresh context) for unbiased assessment. Captures numbers, what went well, didn't go well, lessons, action items.

• senior-engineer-reviewer

  Internal review by Senior Engineer subagent (fresh context). Reviews sprint outcome across Architecture, Code Quality, Tests (with fake test detection), Performance dimensions. Uses provided SE prompt with extended test review for catching fake/weak tests. Required as part of DoD.

• sprint-doctor

  Preflight diagnostic for ai-scrum. Reports hook health, strictness, tracker MCPs, Codex/DeepSeek availability, LSP fallback, test runner, scribe directories. Use before /sprint:start to surface missing tooling and silent fallbacks.

• sprint-edit-plan

  Reviews an imported plan and proposes patches to make it ai-scrum-ready. Detects vague AC, oversized tasks, missing test approach, weak DoR fields, missing PRD/design refs. Used by /sprint:edit-plan.

• sprint-executor

  Core sprint execution loop. Manages task-by-task execution within a sprint with discipline gates: DoR check before task, AC verification after task, hooks active during execution. Used by sprint-orchestrator during sprint execution phase.

• sprint-loop-controller

  State machine for /sprint:loop. Decides next action (pull task, run gate, wait, exit) based on .sprints/active/_state.json + tool-history.log + safety knobs. Used by sprint-loop command.

• sprint-metrics

  Computes cross-cycle metrics over .sprints/archive/. Each metric must trace back to a recorded source (state.json, tool-history.log, retrospective). Used by /sprint:metrics.

• sprint-orchestrator

  Central coordinator for ai-scrum sprint cycles. Manages full lifecycle from /sprint:start through task decomposition, sprint execution, reviews, and HTML report generation. Activated when user invokes /sprint:start or when caller skill declares sprint_required: true.

• sprint-planner

  Groups approved tasks into sprints with goals, increments, DoD. Used by sprint-orchestrator after task-decomposer's plan is approved by decomposition-reviewer.

• sprint-replayer

  Reconstructs sprint cycle timeline from archive. Used by /sprint:replay and /sprint:fork. Walks state.json snapshots, tool-history.log, retrospectives. Produces chronological event stream.

• task-decomposer

  Decomposes high-level scope into atomic tasks with AC, DoR fields, test approaches, and dependencies. Operates in two modes: greenfield (from scope text) or enrichment (using imported external plan as input). Always applies ai-scrum's decomposition algorithm regardless of source - external plans are

• test-first-enforcer

  Enforces test-before-code discipline. Hooks block production code edits without preceding test changes in same task. Used by pre-tool-use-test-first.sh hook for hard enforcement, and consulted by dod-validator for compliance verification.

• tracker-integration

  Detects issue tracker MCPs (Jira, Linear, GitHub Issues, Shortcut). Asks user once per project for opt-in. Syncs sprint tasks as tracker tickets. Updates ticket status as sprint progresses. Adds review findings as comments. Falls back to local-only if no MCP.

Slash commands (23)

• /ai-scrum:sprint-cancel

  Cancel active sprint cycle with mandatory reason. Generates partial retrospective. Cannot be undone.

• /ai-scrum:sprint-checkin

  Mark a step as done in active sprint. checkin-validator verifies honesty by cross-checking against tool history. Required for execution discipline - hooks may block work if no recent checkin.

• /ai-scrum:sprint-codex-status

  Diagnostic for external review pipeline. Reports Codex CLI availability, OpenRouter/DeepSeek fallback presence, last verdict source, geo-block detection.

• /ai-scrum:sprint-complete

  Attempt to close current sprint. Triggers DoD validation gate, internal SE review, external Codex review (if available), retrospective. HARD GATE - blocks if DoD fails.

• /ai-scrum:sprint-doctor

  Preflight diagnostics before /sprint:start. Reports hook status, strictness, tracker MCP, Codex/DeepSeek availability, LSP, test runner, and missing tooling. Returns green/yellow/red verdict.

• /ai-scrum:sprint-dod-customize

  Interactive wizard to customize project DoD. Adds custom criteria to default DoD. Cannot remove mandatory baked-in criteria. Saves to .sprints/config.yaml.

• /ai-scrum:sprint-edit-plan

  Review and patch an imported plan before /sprint:start locks it. Surfaces vague AC, oversized tasks, missing test approach, weak DoR. Saves cleaned plan.

• /ai-scrum:sprint-fail

  Mark a step as failed with mandatory reason. Fallback can be specified via /sprint:fallback.

• /ai-scrum:sprint-fallback

  Replace failed step with alternative approach. Documents the alternative for SE review and retrospective.

• /ai-scrum:sprint-find-plans

  Search project for existing planning artifacts (superpowers plans, PRDs, ADRs, manual markdown plans). Suggests using found plans as input to /sprint:start. Use when you suspect a plan exists but unsure of path, or when continuing work from another plugin.

• /ai-scrum:sprint-fork

  Fork an archived sprint cycle into a new active cycle, preserving original scope/PRD/plan but resetting state. Useful for what-if reruns.

• /ai-scrum:sprint-from-query

  Start sprint(s) from a tracker query (JQL / Linear filter / GitHub search). Bulk variant of /sprint:from-ticket.

• /ai-scrum:sprint-from-ticket

  Start a sprint from an existing tracker ticket (Jira/Linear/GitHub Issues). Pulls title, body, comments, AC, labels via tracker MCP, then runs decomposition.

• /ai-scrum:sprint-history

  Show past sprint cycles from .sprints/archive/. Filter by task name or limit to last N cycles.

• /ai-scrum:sprint-html

  Regenerate HTML report for a sprint cycle (active or archived). Useful after design system changes or template updates.

• /ai-scrum:sprint-loop

  Sprint-aware autonomous loop. Wraps /loop with cadence sensitive to gate state. Auto-pulls next task, sleeps during DoR/DoD validation, exits on cycle complete.

• /ai-scrum:sprint-metrics

  Aggregate metrics across sprint cycles in .sprints/archive/. Reports estimate accuracy, cycle time, gate blocks, test exemptions, force closes, review findings, replan rate. Outputs JSON + HTML dashboard.

• /ai-scrum:sprint-replan

  Re-do task decomposition. Only allowed before Sprint 1 starts execution. Use when initial decomposition proves wrong.

• /ai-scrum:sprint-replay

  Replay an archived sprint cycle. Walks through decisions, gate events, and reviews chronologically. Read-only.

• /ai-scrum:sprint-skip

  Mark a step as skipped with mandatory justification. Used for steps not applicable in current context (e.g., depth=quick skips mutation testing).

• /ai-scrum:sprint-start

  Initialize a sprint cycle for a development task. Triggers context investigation, decomposition with review, and sprint execution with full discipline gates. Use this for any methodological work. Optionally accepts --plan=<path> to import an existing plan from external plugin (superpowers, prd-write

• /ai-scrum:sprint-status

  Show current state of active sprint cycle: which sprint, which task, progress on DoD criteria, recent checkin history.

• /ai-scrum:sprint-test-exempt

  Mark current task as test-exempt with mandatory reason. Allows production code changes без preceding test. Use ONLY for legitimate cases (generated code, configs, migrations). Logged for SE review attention.

Subagents (5)

• codex-bridge

  Bridge to codex-plugin-cc для external review. Sanitizes context, invokes Codex CLI с read-only sandbox, parses results. Falls back to DeepSeek via OpenRouter if Codex unavailable and fallback enabled.

• decomposition-reviewer

  Independent reviewer of task decomposition (fresh context, не biased). Reviews decomposition for atomicity, completeness, dependencies, AC quality. Returns APPROVED or REVISE с specific issues.

• handoff-keeper

  Subagent (fresh context) that updates and validates HANDOFF.md. Ensures sprint outcomes are properly captured. Detects external edits and proposes merge instead of overwrite. Required as part of sprint DoD.

• retrospective-author

  Subagent (fresh context) that writes honest retrospectives. Reduces bias toward defending own work. Cross-checks numbers against execution log, flags vague language, ensures action items are specific. Used at sprint end and task cycle end.

• senior-engineer

  Internal Senior Engineer reviewer subagent (fresh context). Reviews sprint outcomes across Architecture, Code Quality, Tests (extended for fake tests), Performance dimensions. Uses provided SE prompt. Required as part of DoD.

Hooks (4)

• PreToolUse

  ${CLAUDE_PLUGIN_ROOT}/hooks/pre-tool-use-checkin.sh

• PreToolUse

  ${CLAUDE_PLUGIN_ROOT}/hooks/pre-tool-use-test-first.sh

• PostToolUse

  ${CLAUDE_PLUGIN_ROOT}/hooks/post-tool-use-tracker.sh

• Stop

  ${CLAUDE_PLUGIN_ROOT}/hooks/stop-validation-gate.sh

Similar plugins