菜单
Investigate IntexuraOS code-task executions when the user shares a `dev.intexuraos.cloud` or `intexuraos.cloud` code-task URL, or provides a `task_*` ID and asks to debug, investigate, or understand what happened.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Generate professional documentation for IntexuraOS services (apps, workers, packages). Produces 5 doc files per service, 3 per package, plus aggregated site content and cross-validation reports. Supports interactive, autonomous, and team-of-agents modes. Use when documenting services, generating docs, updating documentation, or validating doc consistency.
Orchestrate a 6-phase release workflow with automated service documentation, high-level docs updates, README "What's New" section, website improvements, and semantic versioning. Use when preparing a new release.
Sentry issue triage and investigation with automatic cross-linking. Handles issue investigation, AI analysis via Seer, and integration with Linear for tracking fixes. Use when triaging Sentry errors, investigating sentry.io URLs, or creating Linear issues from Sentry.
Prepare an IntexuraOS release by collecting release data, triaging changes, updating docs, versions, changelog, website release highlights, and GitHub release notes.
Use when the user says "commit push", "commit and push", "push a PR", or asks Codex to finalize IntexuraOS changes by committing, pushing, and opening a pull request against development.
Use when user pastes a dev.intexuraos.cloud or intexuraos.cloud code-task URL and asks to debug, investigate, or understand what went wrong. Also use when user mentions a code task ID (task_*) with investigation intent.
| name | debug-code-task |
| description | Investigate IntexuraOS code-task executions when the user shares a `dev.intexuraos.cloud` or `intexuraos.cloud` code-task URL, or provides a `task_*` ID and asks to debug, investigate, or understand what happened. |
Use this skill for code-task investigations backed by Firestore data.
https://dev.intexuraos.cloud/#/code-tasks/task_*https://intexuraos.cloud/#/code-tasks/task_*task_* plus intent such as debug, investigate, or what went wrongtask_* input.dev.intexuraos.cloud = devintexuraos.cloud without dev. = produname -n..codex/skills/debug-code-task/scripts/fetch-task.sh, which resolves the local repo and runs the Firestore fetcher from the checked-out repo.Run:
.codex/skills/debug-code-task/scripts/fetch-task.sh <taskId>
The wrapper uses the checked-out repo for firebase-admin module resolution and expects credentials at ~/.config/gcloud/sa-key.json.
Summarize these fields for the user when present:
idstatuslinearIssueIdworkerLocationagentTyperesult.summaryerrorRun:
.codex/skills/debug-code-task/scripts/fetch-task.sh <taskId> --logs-only
Use --logs if you want the task document and logs in one call. Logs are ordered by sequence number and printed as [sequence] message.
Present log evidence directly. Do not infer a root cause unless the logs or task document support it.
Only do this when Firestore logs are not enough or the user asks.
workerLocation from the task document first.home-dev and you are elsewhere, use ssh home-dev.home-dev:journalctl -u intexuraos-orchestrator@pbuchman --since "<time>" --until "<time>"
docker logs claude-worker-<taskId>
Derive the time window from the task document timestamps and pad it by about one minute on each side.
curl or web-fetch the code-task SPA URL./tmp.node -e for commands that may include !..codex/skills/debug-code-task/scripts/fetch-task.sh as the canonical wrapper path.