Browse Skills

Gstack Review performs a pre-merge review of the current diff to catch regressions, security issues, missing tests, and cleanup problems. Use it before shipping or asking whether a change is safe.

Provides fixes for common OpenChatCut tool call issues, including edit_item shape errors, timeline overlaps, workspace path restrictions, and motion graphic requirements.

Provides verification levels and tools for confirming agent edits in the OpenChatCut project and editor. Includes static checks, runtime editor observation, and desktop app validation.

Agent Qa Authoring helps create, edit, validate, and run agent-qa tests, suites, and hooks. Use it when you need canonical IDs, schema-checked config, or MCP-based authoring and run workflows.

Agent Qa Debug Fix uses MCP evidence, logs, artifacts, and local code changes to diagnose a failed agent-qa run and patch the root cause. Use it after a run fails and you need to verify the fix with the narrowest relevant test or hook.

Agent QA Result Triage investigates failed agent-qa runs by checking run status, artifacts, logs, and step output. Use it to classify failures, compare recent runs, and report evidence-backed next actions.

Debug Live Issue helps investigate production-like failures with evidence-based steps. Use it for regressions, flaky behavior, workflow breaks, or data inconsistencies across hooks, API, DB, websocket, and UI.

File Headers enforces copyright and file-overview headers at the top of applicable source files. Use it when creating or editing JS, TS, Python, shell, or CSS files that must stay compliant.

Hook Inventory catalogs Claude Code hooks across user, project, and project-local settings and checks them for network egress, arbitrary execution, and missing timeouts. Use it when auditing hook safety or validating hook scripts under ~/.claude/hooks.

Release Guard runs release-readiness checks for docs, scripts, verification coverage, and operational safety before merge or release. Use it to compare command consistency, architecture docs, and safety controls against the current code.

Repo Onboarding helps you understand a codebase before making changes. Use it to find the relevant layer, map ownership, choose files, and pick verification commands.

Closed-Loop Execute runs a V-model pipeline from plan to build, component verification, integration verification, and acceptance checks. Use it for tasks that need traceable evidence tied to acceptance criteria and post-condition validation.