csark0812
- 16 skills
- 0 followers
- 6 hours ago last updated
- ▌ Skeleton · csark0812Agent ops manual for skeleton-enabled repos — init, catalog, audit, optional customize hooks, and toolbox skill overrides. Use when editing skeleton.toml / .skeleton/, syncing toolbox skills, or running skeleton CLI.
- ▌
- ▌
- ▌ Tdd · csark0812 bundleTest-first build at agreed public seams, one red-green slice at a time. Process skill. Not repro-first hard-bug work without a seam, a throwaway spike, or find-only hunch settlement.
- ▌ Grill · csark0812 bundleClarify fuzzy intent and pressure-test design choices through focused dialogue before implementation. Use when the user wants to shape an idea, resolve consequential tradeoffs, or says to grill or crystallize it. Not written-artifact critique, bug diagnosis, or implementation.
- ▌ Probe · csark0812 bundleNarrow a hunch or hard bug through evidence and a pragmatic-STE verdict. Fix only with explicit authority and an on-demand failing signal. Not written-artifact critique, open ideation, or greenfield test-first work.
- ▌ Council · csark0812 bundleMulti-agent orchestrator that creates task-specific personas, selects a useful interaction pattern, runs real independent members, and synthesizes one clear answer. Use when the user attaches or names council for multi-perspective depth on a concrete task. Not single-pass critique, cross-session handoff, or repeated pass loops.
- ▌ Handoff · csark0812 bundleAgent-to-agent cross-session transfer — pointers not bodies. Channel prompt (user /handoff) or artifact (model-invoked subagent). Pack and Goal are open vocabulary. Orchestrator for continuing work in a fresh chat. Not same-session continuation without a new chat.
- ▌ Prototype · csark0812 bundleCreate a throwaway logic, state, or UI artifact that answers one design question. Not production test-first implementation or repro-first debugging.
- ▌ Code Review · csark0812 bundleReview code through a user-named surface and lens. Use for focused risk checks, ordinary code review, prior-finding closure, or a strict code-quality merge gate. Read-only unless the user separately asks for fixes. Not hunch settlement, written-artifact critique, or multi-agent orchestration.
- ▌ Domain Model · csark0812 bundlePersist glossary entries and ADRs after a term or decision is ready. Process skill. Not open design dialogue.
- ▌ Second Opinion · csark0812 bundleAdversarial critique of a written artifact through task-specific lenses and claim anchoring. Single-agent by default; compatible with user-attached multi-agent orchestration. Not dialogue-only design or find-only hunch settlement.
- ▌ Refine Agent Work · csark0812Judge agent-created work against the developer's preferences, record ownership decisions, and refine accepted changes in bounded proven slices. Not a general walkthrough or a pre-decided structural refactor.
- ▌ Refactor Companion · csark0812 bundlePreserve a developer's target design through evidence-led, proven refactor slices. Use for existing-code replacement, simplification, migration, cutover, ownership change, or deliberate removal of an old shape. Not ordinary feature work, analysis-only review, or autonomous Git operations.
- ▌ Review Walkthrough · csark0812 bundleBind a code change and explain it through the smallest useful causal story. Use when a user wants to understand staged, working-tree, commit, branch, pull-request, or scoped-path changes before deciding what to do next. Read-only; not formal review or repair.
- ▌ Remote Ambient Probe · csark0812Validation-only fixture — proves whether ambient refs via GitHub URLs work at agent runtime. Not a shipped toolbox skill.