@Chachamaru127
Going all-in on the AI business from April 2025 OSS release: 3k Stars AI adoption starting with "I want" AX Support | Tokenomics | AISDR |
25 skills1 repository 3.2K
Create a standalone HTML planning brief before implementation. It summarizes the request, options, risks, acceptance criteria, and readiness evidence, using relevant decisions and past plans from the current project.
Prepare and run a GitHub release for projects using Keep a Changelog. It detects a version bump, drafts release notes, then after one confirmation updates version files, merges the release to the default branch, tags it, and verifies publication by the project workflow.
Run unfinished Plans.md tasks over repeated wake-ups, restarting with fresh context each time. The loop executes and reviews one task per cycle, records progress, and stops when work is complete, a cycle limit is reached, or escalation is needed.
Send a question, investigation, or design discussion to cursor-agent in read-only ask mode. It returns a second opinion or evidence-based findings that the host agent summarizes, without making file changes or running implementation work.
Set up or diagnose Harness project configuration, including project initialization, CI, Codex CLI, memory, agents, and skill mirrors. Choose a focused setup area, follow its reference guidance, and check the resulting configuration where needed.
Turn task statuses in Plans.md into a standalone progress dashboard. It shows task counts and completion percentage, elapsed and estimated time, and reported costs, with support for regenerating the HTML snapshot.
Execute implementation tasks listed in Plans.md, from one selected task to a larger parallel run. The workflow chooses an execution mode based on task count or explicit options, then validates and reviews the results before reporting completion.
Send verified updates or handoffs to an agent in another session. The workflow checks the active session list to identify the correct team and agent, then formats a message with relevant scope, evidence, and remaining decisions.
Delegate one implementation or file editing task to Cursor Composer in a dedicated Git worktree. Review its changes, then cherry-pick the reviewed commit into the current branch and update the matching plan entry when present.
When someone corrects the agent’s Japanese wording or style, capture the before-and-after as a pending writing rule proposal for later human review. The skill does not approve proposals or change the active rules dictionary.
Run planned software tasks as a coordinated team workflow. It assigns ready tasks to isolated workers, performs per-task and integrated reviews, incorporates approved changes, and reports progress and completion. Worker execution can use the selected backend, including optional Cursor delegation.
Run planned work through a team workflow: delegate implementation to workers, coordinate independent review and revisions, then report the outcome. Select a task range or backend, or run all unfinished tasks in one command.
Review changes that integrate Claude Code or Codex updates. Classify each new Feature Table entry as implemented, inherited upstream, or tracked in Plans, and flag entries that are only documented without supporting work.
Review code changes, plans, or project scope from several angles, including security, quality, specification fit, and regression risk. Get a verdict supported by inspected evidence, plus findings and next steps when changes are needed.
Use the agent-browser CLI to navigate web pages, inspect interactive elements, fill forms, capture screenshots, and collect page information. It helps check web app behavior through browser actions and snapshots.
Diagnose failed CI runs by reviewing logs, reproduction details, specifications, and recent commits. Identify whether the cause lies in implementation, tests, or the environment, then make authorized fixes without hiding failures by skipping checks.
Check whether the Cursor package and agent are ready for Claude Code Harness, and see which implementation backend is resolved. When requested, set Cursor as the user or project default, or remove that setting.
Review a delivered task against acceptance criteria saved during planning. Build a standalone HTML demo that marks each criterion verified or unverified, shows supporting evidence, and recommends whether to ship, wait, or reject.
Turn ideas and requirements into scoped, checkable project tasks, keeping the product specification and Plans.md aligned. Add tasks, update status markers, and compare recorded progress with implementation evidence to identify synchronization gaps.
Compare task markers in Plans.md with Git changes and recent edit traces to find progress drift. Get a status summary and suggested updates, save a timestamped snapshot, or review completed work for patterns and lessons.
Start a real background runner for long-running Codex work. It groups ready tasks from Plans.md by dependencies, executes each batch through Breezing by default, and lets you check status or stop the loop.
Search and manage shared project memory across Claude Code, Codex, and OpenCode. Find past decisions, sync or migrate records, and propose promoting important learnings into a single source of truth while preserving their sources and reasoning.
Clean up project planning files, session logs, old logs, and state files by archiving completed work, splitting long logs, or trimming stored data. Preview proposed actions with dry-run and receive a before-and-after summary.
Analyze orchestration logs and the Judgment Ledger to identify recurring failures and produce confidence-scored failure-rule.v1 proposals. The skill works read-only; any proposed addition to patterns.md or decisions.md requires human review and is never applied automatically.
Is this your account? Link GitHub to bring it onto your creator page (Japanese)