Claude Skills by castbox

Execute one bounded task-free change through AI-owned checkout suitability, editing, targeted validation, and risk-evolution review.

Finalize a reviewed Trellis task through one semantic finalization loop, one deterministic transaction engine, and six typed exits.

Persist Completion and Closure into one reviewed task archive and terminal bookkeeping result.

Maintain a repository Architecture Baseline through semantic bootstrap, task impact synchronization, promotion, and repair profiles.

Maintain one traceable Requirements, Design, and Test repository authority through semantic profiles.

Merge one reviewed active-task Delivery PR through a semantic live gate, exact merge-commit trailers, and terminal recovery.

Merge one Ready task PR through a semantic live gate, or request an independent read-only review of an exact completed archive.

Publish one reviewed active-task Delivery through a semantic gate and an idempotent push, PR binding, metadata, and Ready transaction.

Qualify one profile-specific candidate set against current normal-scenario authority before a Guru Team caller may promote it into scope, tests, findings, implementation, or publication blocking.

Qualify whether a proposed implementation mechanism can carry the required business authority without delegating it to OS, kernel, process, or descriptor primitives.

Reactivate one normally finished archived task while preserving its identity and routing the fresh work.

Reconcile one exact active-task and evolved-base pair through a semantic impact review, controlled integration evidence, and six typed exits.

Restore one archived Trellis task to Phase 2 after a scoped merge blocker through fresh identity checks and an idempotent local transaction.

Review committed task content, bounded base continuity, or a completed archive through independent semantic review.

Review whether a current change request is one independently deliverable unit through linked prerequisite evidence, an AI readiness gate, and five typed exits.

Review controlled contract wording through fixed change-request, planning-artifact, or explicit-path scope, semantic revision and classification, deterministic evidence, and typed exits.

Review all current task Delivery and evidence facts and select one bounded completion route.

Review one approved active-task Delivery slice through a fresh semantic gate and five minimal typed exits.

Review task publication readiness through ten semantic dimensions, ordinary metadata revision or read-only archived review, and four typed exits.

Select standard Guru Team intake or a bounded task-free current-checkout route through one AI-owned semantic decision.

Resolve, safely synchronize, and deterministically validate the selected Git base before Guru Team reads repository task context or starts intake side effects.

Verify a Guru Team source checkout through a clean standalone throwaway installation, semantic adequacy review, and two terminal exits.

Prepare and publish an official castbox/guru-trellis version through the repository-private two-stage release contract without joining the public Guru Team extension inventory.

Guides collaborative requirements discovery before implementation. Creates task directory, seeds PRD, asks high-value questions one at a time, researches technical choices, and converges on MVP scope. Use when requirements are unclear, there are multiple valid approaches, or the user describes a new feature or complex task.

Understand and customize the local Trellis architecture inside a user project. Use when modifying .trellis plus platform hooks, settings, agents, skills, commands, prompts, workflows, the channel runtime (trellis channel), bundled runtime agents under .trellis/agents/, selectable workflow templates, registry-backed spec refresh, cross-session memory (trellis mem) generated by trellis init, or AI-facing bundled skills (trellis-channel, trellis-session-insight, trellis-spec-bootstrap) and bundl...

Reach into past AI conversation history through the `trellis mem` CLI. Use whenever the user asks 'how did we solve X last time', 'have we discussed this before', 'what was the decision on X', 'remind me what we did in this task', '上次怎么解的', '之前讨论过吗', '想起一段对话', or when starting a brainstorm that overlaps prior work, debugging a familiar bug, continuing a task across sessions, or doing a finish-work review. Returns raw past dialogue; decide for the moment whether to update spec, append to task ...

Bootstrap project-specific Trellis coding specs with a platform-neutral single-agent workflow. Use when creating or refreshing .trellis/spec guidelines, analyzing a codebase with ABCoder or source inspection, decomposing package/layer spec work, and writing real codebase-backed spec docs without placeholder text.

Approve current task planning through one AI-owned semantic review, compact private evidence, and four minimal typed exits.

Bind one current session to an exact task identity, rebind after loss, switch between independent tasks, and resume with lifecycle freshness.

Bootstrap and repair the repository Requirements/Design/Test and Architecture Baseline SSOTs with a minimal .trellis/spec projection.

Check the complete current task through one AI-owned semantic round, compact private evidence, and four minimal typed exits.

Clarify initial requirements, active-task scope changes, or explicit review requests through an evidence-first semantic loop with dialogue-local choices and typed exits.

Remove only the current Finish-owned branch, worktree and runtime resources after terminal success.

Apply the exact source Issue disposition after Completion approval and return a recoverable closure result.

Create a reviewed Trellis task work commit after Phase 2, including finding-fix and revision commits, with exact staging and fresh evidence.

Create or exactly reuse a reviewed issue workspace and Trellis task through an AI-owned semantic gate, ephemeral scoped confirmation, deterministic execution, and four typed exits.

Discover fresh current and archived change context, run the semantic evidence gate, and hand minimal current identity to Guru Team requirement clarification.

Execute one bounded task-free change through AI-owned checkout suitability, editing, targeted validation, and risk-evolution review.

Finalize a reviewed Trellis task through one semantic finalization loop, one deterministic transaction engine, and six typed exits.

Persist Completion and Closure into one reviewed task archive and terminal bookkeeping result.

Maintain a repository Architecture Baseline through semantic bootstrap, task impact synchronization, promotion, and repair profiles.

Maintain one traceable Requirements, Design, and Test repository authority through semantic profiles.

Merge one reviewed active-task Delivery PR through a semantic live gate, exact merge-commit trailers, and terminal recovery.

Merge one Ready task PR through a semantic live gate, or request an independent read-only review of an exact completed archive.

Publish one reviewed active-task Delivery through a semantic gate and an idempotent push, PR binding, metadata, and Ready transaction.

Qualify one profile-specific candidate set against current normal-scenario authority before a Guru Team caller may promote it into scope, tests, findings, implementation, or publication blocking.

Qualify whether a proposed implementation mechanism can carry the required business authority without delegating it to OS, kernel, process, or descriptor primitives.

Reactivate one normally finished archived task while preserving its identity and routing the fresh work.

Reconcile one exact active-task and evolved-base pair through a semantic impact review, controlled integration evidence, and six typed exits.

Restore one archived Trellis task to Phase 2 after a scoped merge blocker through fresh identity checks and an idempotent local transaction.