my-agent-skills

Autonomously audit native local Codex session JSONL for workflow quality, latency, cost shape, tool use, recurring failures, and eval candidates. Use when the user wants scheduled harness observability or evidence-driven harness improvement; do not use for ordinary task execution, application telemetry, or full transcript export.

Verify an existing build artifact itself: metadata, complete download, checksum or size, archive integrity, unpacked contents, and source or lock provenance. Use when artifact contents or completeness must be proven; do not use to trigger builds, releases, deploys, or health checks.

Diagnose the live remote-access layer: SSH resolution, reachability, Tailscale, login-shell PATH, service state, package-manager locks, and remote CLI auth. Use when a named remote host or service is unreachable or unhealthy; do not use for deploy planning or ordinary local repo work.

Establish the current repo, instruction, branch, dirty-worktree, and linked live-object facts before non-trivial work. Use once at the start of repo work when current state or scope is unclear; do not stack it onto every later workflow phase.

Carry already-scoped code changes through verification, commit, push, PR/MR, authorized merge, and authorized cleanup. Use when the requested outcome includes delivery actions; do not use for review-only work, initial repo discovery, CI diagnosis, or deploy execution.

GitLab CLI workflows with `glab` for issues, merge requests, CI/CD, and GitLab API operations. Use when the user needs GitLab command-line automation, issue or MR triage, pipeline checks, API queries, threaded discussion replies, release checks, or work item conversion and hierarchy updates. Prefer `glab api` for reliable scripted flows, while keeping direct CLI commands and local helper scripts available for common MR and pipeline tasks.

Audit and fix repository-level agent scope and permission overreach in prompt or settings files. Use when an agent can execute unrelated commands, escape its stated task, or has contradictory allow/deny rules; do not use for general security review or ordinary application authorization bugs.

Isolate the first actionable failure in an already-red GitHub Actions or GitLab CI run. Use when a concrete PR, MR, workflow, job, or pipeline is failing; do not use for general repo preflight, local test selection, or healthy deploy verification.

Turn an explicitly requested goal or fuzzy objective into a concrete, measurable outcome and create it with the goal tool when asked. Use only for goal definition or refinement; do not force goal creation onto ordinary implementation, planning, checkpoint, or handoff work.

Resolve target, workflow inputs, exact source SHA, authority, and live verification before release, deploy, packaging, or workflow dispatch. Use for operations that may mutate a release or environment; use artifact-verify instead when only inspecting artifact contents.

查询 Tianbot 官方文档并回答 ROS2GO、TOM、Tianracer、Tianrover、Tianbot mini、ROSECHO 等产品的使用、恢复、兼容性、升级、备份、规格和故障排查问题。仅在 Tianbot 产品问题上使用，并实时核对官方页面。

Use the bundled Exa API wrapper whenever a task needs live web research, current external sources, technical documentation lookup, code examples, API usage context, or company background research with source URLs, especially when local files are not enough and an MCP server is unnecessary.

使用 RapidOCR 对本地图片做 OCR。当用户要从 PNG/JPG/JPEG/WEBP/BMP/TIFF 图片中提取文字、返回结构化 JSON、标记低置信度或疑似金额数字，或把表格截图粗略还原成 Markdown 表格时使用；只处理本地文件，不下载远程 URL。

使用 OpenCLI 处理基于浏览器登录态的网页渠道与网页读取任务。只要用户是在网站内查资料、读取网页内容、导出网页正文、抓取登录后页面、读取前端渲染结果，或要使用 OpenCLI 已支持的网站/渠道命令（如 doubao、weixin、zhihu、youtube、bilibili），就优先使用这个 skill；先用站点/渠道专用命令，`web read` 仅作 fallback。不适用于桌面 App 控制、页面探索、录制/生成适配器、通用 UI 自动化或带副作用的写操作。