observal-advanced

Recovers Observal session ingestion, manages CLI upgrades, downgrades and rollback, and performs explicit local Agent fallback when the server is unavailable. Use when the user asks to reconcile missed sessions, repair CLI version state, or continue locally after a confirmed connection or configuration failure.

跳到安装

仓库: Observal/Observal
最近来源活动: 2026年8月16日 21:32
检测到的 SKILL.md 语言: 英语
星标: 2,351
分支: 471

安装方式

默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。

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。

直接命令不会经过审查 Prompt；运行前请先检查来源。

npx skills add https://github.com/Observal/Observal --skill observal-advanced

命令会保持在同一行。复制前请横向滚动并检查完整内容。

想先保存到本地？可下载 SkillsMP 当前能够提供的文件。

检查来源文件

决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。

文件资源管理器

2 个文件

SKILL.md

来源说明 · 只读预览

返回搜索返回安装

observal-admin

Administers Observal users, settings, diagnostics, review queues, security events, audit logs, SAML, SCIM, local server services, upgrades, rollback, and database migrations. Use when the user needs privileged governance, submission decisions, identity configuration, security investigation, or server operations.

2026-08-16

observal-agents

Creates, authors, validates, publishes, updates, versions, pulls, archives, restores, transfers, and manages co-authors for Observal Agents. Use when the user wants to build or install an Agent, change an Agent definition, publish a draft, release a version, or manage Agent ownership.

2026-08-16

observal-ops

Inspects Observal traces, sessions, rankings, feedback, telemetry health, logs, and Agent insight reports. Use when the user wants operational evidence, current activity, telemetry diagnosis, ratings, report generation, regression analysis, or recommendations grounded in Agent usage.

2026-08-16

查看这个仓库的全部 Skills

Recovering Observal

Execution contract

Execute commands with a 60 second timeout unless the operation documents a longer wait.

Use machine output by default: add --output json whenever supported. Parse list results from items and pagination fields.

Run --help before acting when a path or flag is uncertain.

Use dry run before reconciliation when scope or session volume is uncertain.

Supply documented force flags so version operations never prompt.

Verify cursor, outbox, installed version, checksum, and rollback state after recovery operations.

Fail openly. Never hide an unavailable server behind automatic local writes.

Never retry reconciliation, version changes, or fallback writes without checking resulting state.

Read Recovery workflows completely before executing.

Decision rules

Healthy telemetry does not need routine reconciliation.

Reconcile repairs missed local session delivery. It does not replace hook or extension installation.

Upgrade, downgrade, and rollback are distinct requests. Do not substitute one for another.

Local fallback is allowed only after the CLI explicitly reports Connection failed or Not configured, and only when the user still wants local files written.

Local fallback creates harness-native Agent files. It does not publish Registry state and must be reported as local-only.

Never invent telemetry environment variables or wrappers.

Completion

Report sessions discovered, queued, skipped, or failed for reconciliation; old and new versions for CLI changes; or exact local paths for fallback. Include unresolved warnings and the command needed once the server is reachable.