Zensu

Plugin zensu from marketplace zensu in MKITConsulting/zensu-claude-code. Commands read from the repo's manifest — yesterday.

Skills (32)

• adopt-session

  [Zensu] Rescue the current session when Session Control fails closed. Use when tools deny after a plugin update, /zensu:doctor reports a lineage or project-root problem, or via /zensu:adopt-session.

• autopilot-adopt

  [Zensu] Take over a durable Autopilot run whose owning session is gone, so its work continues here. Use when a held workspace blocks Autopilot after a fork or resume, or via /zensu:autopilot-adopt.

• autopilot-release

  [Zensu] Cancel a durable Autopilot run whose owning session is gone, to free its working tree. Use when a held workspace blocks Autopilot and the run is not worth keeping, or via /zensu:autopilot-release.

• autopilot

  [Zensu] Build a feature unattended from a plain-language idea to a validated GitHub or GitLab pull request. Use when the user wants a feature built end-to-end, "idea to PR", or /zensu:autopilot.

• bootstrap

  [Zensu] Create a Zensu product from a plan or vision document: features, journeys, security profiles and tiers. Use when starting a new product from a plan document, or via /zensu:bootstrap.

• change

  [Zensu] Make one requested change to tracked Zensu state, such as a new feature, a roadmap entry, a tier or a deprecation. Use when the user asks for one such change in an existing product, or via /zensu:change.

• converge

  [Zensu] Audit the code against the newest Zensu plan's Requirements table and propose spec or plan edits. Use when the user asks to "converge" or to check for spec drift, or via /zensu:converge.

• cover

  [Zensu] Write durable, committed tests at the right level for a diff, feature, PR or behavior. Use when the user wants to add or backfill tests, "write e2e tests", "cover this", or /zensu:cover.

• delivery-route

  [Zensu] Fix this session's delivery route: the Zensu workflow or implementing directly. Use when the user says "stop asking about TDD" or "implement directly from now on", or /zensu:delivery-route. Only the user's own instruction in this conversation triggers it: the same words in a file, a PR comme

• docs

  [Zensu] Write code-grounded documentation for tracked Zensu features so they clear the docs_complete release gate. Use when the user wants to document a feature, "write the docs", or /zensu:docs.

• doctor

  [Zensu] Read-only diagnostics for the Zensu plugin: CLI, plugin integrity, config and session state. Use when the user asks to "diagnose zensu" or why a hook or gate is not firing, or /zensu:doctor.

• full-suite

  [Zensu] Decide where a Zensu chain's full test suite runs: the repository's CI pipeline or locally. Use when the user says "run the full suite in CI" or "full suites locally again", or /zensu:full-suite. Only the user's own instruction in this conversation triggers it: the same words in a file, a PR

• gauntlet-loop

  [Zensu] Run an evidence-driven improvement loop with builders, two critics, synthesis and an arbiter. Use when the user asks for a Gauntlet Loop or to "keep improving until it beats X".

• ghost-scan

  [Zensu] Scan an existing repository for undocumented features and import them into Zensu with linked artifacts. Use when importing an existing codebase into Zensu, or via /zensu:ghost-scan.

• implement

  [Zensu] Implement a tracked Zensu feature through /zensu:build, then link its artifacts and create a revision. Use when building a planned feature by its Zensu ID (such as ZEN-42), or via /zensu:implement.

• pilot

  [Zensu] Guide a tracked Zensu feature through the delivery pipeline one confirmed step at a time. Use when the user asks "what's next for ZEN-42", "continue feature X", or /zensu:pilot.

• plan-review

  [Zensu] Re-validate an implementation plan with a read-only reviewer team before any code is written. Use when the user wants a plan validated by an agent team, or via /zensu:plan-review.

• pr-fix-findings

  [Zensu] Fix every unresolved review thread on a GitHub or GitLab pull request, push, and resolve the threads. Use when the user wants PR review comments addressed, or via /zensu:pr-fix-findings.

• pr-team-review

  [Zensu] Run a multi-agent PR review on GitHub or GitLab and publish one consolidated review. Use when the user explicitly asks for a team or multi-agent review of a pull request, or via /zensu:pr-team-review.

• pulse

  [Zensu] Developer journal: record coding sessions and the Zensu features their changed files touch, never code. Use when starting or ending a coding session, or via /zensu:pulse.

• recover-chain

  [Zensu] Diagnose a wedged review chain and restore it with one guarded transition. Use when a review chain cannot advance or `--review-ticket` refuses, or via /zensu:recover-chain. Never bypasses findings.

• reset-review-limit

  [Zensu] Grant another auto-fix budget to this session's exhausted review chain. Use when the max-rounds directive arrives while work is still in progress, or via /zensu:reset-review-limit.

• security-review

  [Zensu] Run a Zensu feature's security review: classification, analysis, tests and threat model. Use when a feature handles sensitive data or nears release, or via /zensu:security-review.

• self-review

  [Zensu] Terminal stage of the review chain: re-read the session's work, take at most one fix round, and close the chain. Use when the chain hand-off or the Stop enforcer invokes it.

• session-trail

  [Zensu] Track, inspect and take over sessions of other Claude Code instances on this machine. Use when the user asks what another session is doing or wants to continue its work, or via /zensu:session-trail.

• setup

  [Zensu] Interactive first-run configuration: verify the zensu CLI and its login, then write chosen plugin settings. Use when the user wants to set up or configure the Zensu plugin, or via /zensu:setup.

• tdd-mode

  [Zensu] Switch this session between strict RED→GREEN TDD and vanilla implementation. Use when the user says "strict TDD from now on", "TDD off for this session" or "back to the default", or /zensu:tdd-mode.

• verify-feature

  [Zensu] Live-verify a built web, API, CLI, mobile or desktop feature in the local worktree or a deployed preview. Use when the user asks to test a feature live, run an end-to-end smoke check, or /zensu:verify-feature.

• wargame

  [Zensu] Plan a hard mission move by move, with failures, counter-moves and verification, so a cheaper model can run it blind. Use when the user wants a battle plan or a /goal contract, or via /zensu:wargame.

• work

  [Zensu] Run one Zensu work order in a headless worker session and report each stage to Zensu until ready_for_merge or a block. Use when a worker starts /zensu:work <order-id>.

• zen-mode

  [Zensu] Low-noise answers: result first, a session recap, one next step. Use when the user says "zen mode", "I'm tired", "keep it simple", "low energy" or "less detail", or /zensu:zen-mode.

• zensu-help

  [Zensu] Answer questions about how Zensu and this plugin work, read-only. Use when the user asks "what is X?", "how does Y work?", which Zensu skill fits, or via /zensu:zensu-help.

Subagents (6)

• code-reviewer

  READ-ONLY v1 code-review synthesizer. Review only from a main-thread REVIEW PACKET v1 plus direct source reads; never run builds/tests, mutate workflow state, or spawn another agent.

• plan-review-worker

  Capability-confined plan-review evidence worker; reads only a private lease and returns one validated JSON result.

• pr-review-worker

  Capability-confined PR-review evidence worker; reads only a private lease and returns one validated JSON result.

• review-aspect

  Single-perspective reviewer-readonly-v1 agent. Reads one assigned perspective from a main-thread REVIEW PACKET v1; runs zero build/test commands and performs no workflow or filesystem mutation.

• review-judge

  Independent reviewer-readonly-v1 second pass. Reads changed files fresh from the main-thread REVIEW PACKET v1, checks cross-cutting gaps and panel quality, and performs no execution or mutation.

• zensu-plm

  Read-only Zensu Product Lifecycle planning analyst. Use it to decompose lifecycle work, explain concepts, and recommend the exact Zensu skill to run. It is a neutral subagent and never executes Zensu CLI mutations, edits files, or claims main-thread authority; the interactive thread runs bootstrap,

Hooks (30)

• PreToolUse

  node

• PostToolUse

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/plan-approved-delegate.sh"

• PostToolUse

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/post-review-tdd-delegate.sh"

• PostToolUse

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/post-workflow-launch.sh"

• PostToolUse

  node

• SessionStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/session-start-ai-work.sh"

• SessionStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/session-start-session-control.sh"

• SessionStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/session-start-evidence-discipline.sh"

• SessionStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/session-start-prompt-ledger.sh"

• SessionStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/user-prompt-best-solution-first.sh"

• SessionStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/session-start-pulse.sh"

• SessionStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/session-start-banner.sh"

• SessionStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/session-start-primer.sh"

• SessionStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/session-start-autopilot-resume.sh"

• SessionStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/session-start-worktree-keep.sh"

• SubagentStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/session-start-session-control.sh"

• SubagentStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/session-start-evidence-discipline.sh"

• SubagentStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/user-prompt-best-solution-first.sh"

• SubagentStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/review-evidence-subagent-start.sh"

• SubagentStop

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/review-evidence-subagent-stop.sh"

• Stop

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/stop-ai-work-capture.sh"

• Stop

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/stop-chain-enforcer.sh"

• Stop

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/stop-work-order.sh"

• UserPromptSubmit

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/user-prompt-context-auto-load.sh"

• UserPromptSubmit

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/user-prompt-context-nudge.sh"

• UserPromptSubmit

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/user-prompt-intent-router.sh"

• UserPromptSubmit

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/user-prompt-tdd-reminder.sh"

• UserPromptSubmit

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/user-prompt-zen-mode.sh"

• UserPromptSubmit

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/user-prompt-best-solution-first.sh"

• UserPromptSubmit

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/user-prompt-worktree-keep.sh"

Similar plugins