rollback-planner
Generate rollback plans for proposed changes — defining exactly how to undo every operation safely. Use when the user requests rollback planner or invokes rollback-planner. Do NOT use for unrelated application features or unapproved production changes.
- License 许可
- CC-BY-4.0
- 最近更新
- 2026-11-14
- Stars
- ★ 1
- Forks
- 0
01 前言 Front Matter
- name
- rollback-planner
- license
- CC-BY-4.0
- metadata
- { "author": "AEOS", "version": "1.0.0", "owner_agent": "codenavi-agent" }
- description
- Generate rollback plans for proposed changes — defining exactly how to undo every operation safely. Use when the user requests rollback planner or invokes rollback-planner. Do NOT use for unrelated application features or unapproved production changes.
02 资源文件 1
- 参考文档references/provenance.md533 B
03 SKILL.md 正文
Rollback Planner Skill
Governance: CodENavi Full Workspace v2. Skill of codenavi-agent; inherit AGENTS.md.
Lifecycle: BRIEFING → RECON → PLAN → EXECUTE → VERIFY → DEBRIEF.
Source roles/workers are internal lenses. Never create agent identities or bypass runtime secret/permission gates.
ID: rollback-planner Version: 1.0.0 Owner Agent: codenavi-agent Risk Level: low
Mission
Generate rollback plans for proposed changes — defining exactly how to undo every operation safely.
Scope
- Analyze proposed changes from patch
- Define undo operations for each change
- Generate rollback-plan.md with verification steps
Allowed Actions
filesystem.read— read patch filesfilesystem.write_aeos— write rollback plan to .aeos/patches/{execution_id}/generate_evidence— register rollback plan
Forbidden Actions
- Execute any rollback operation
- Modify real files as part of planning
- Delete any real files
Required Capabilities
- READ_REPOSITORY
- PLAN_ROLLBACK
- GENERATE_REPORT
Required Evidence
- Patch files read (SHA-256)
- Rollback plan generated (SHA-256)
- Verification steps documented
Prompt Contract
- State the objective, target scope, assumptions and constraints before execution.
- Use only evidence-backed facts; mark uncertainty explicitly.
- Route tool access through approved command, MCP or Tool Router paths.
- Redact secrets, credentials, tokens and sensitive values.
- Return facts, assumptions, risks, recommendations, evidence_refs and blocking_conditions when applicable.
- Stop when required evidence, permissions, policy approval or input context is missing.
Quality Gates
- Every proposed change must have a corresponding undo operation
- Rollback must include verification steps
- Rollback plan must be actionable
Output Schema
{
"strategy": "restore_originals + delete_generated",
"operations": [
{"operation": "restore", "target": "src/service.py", "method": "git checkout -- src/service.py"},
{"operation": "delete", "target": ".aeos/patches/ex-123/proposed.patch", "method": "rm"},
{"operation": "verify", "target": "", "method": "git diff --stat && run tests"}
],
"rollback_conditions": ["regression", "test_failure", "approval_revoked"]
}
Blocking Conditions
- Any change without undo operation
- Rollback plan is empty
- Verification step is missing
Examples of Allowed Behavior
- Create rollback-plan.md for a patch that modifies src/service.py
- Define git checkout commands for each modified file
Examples of Forbidden Behavior
- Executing the rollback plan
- Modifying source files as part of planning
Provenance and source freshness: provenance.
04 同仓库更多 Skills
workspace-brain
Operate the CodENavi workspace brain MCP and CLI — cloud DeepSeek RAG, disk corpus at WORKSPACE_BRAIN_ROOT, Jev URL curation, and event logging for Cursor/Codex/OpenCode/Claude. Use when ingesting repo knowledge, asking the brain, mapping web seeds, or recording AI-driven file changes. Do NOT use for local Ollama RAG or storing API keys in git.
workbench-profile-resolver
Provide enterprise-grade capability for `workbench-profile-resolver` inside AEOS v1.1. Use when the user requests workbench profile resolver or invokes workbench-profile-resolver. Do NOT use for unrelated tools or unapproved remote writes.
wordpress-expert
Staff-level WordPress engineering super-skill for one-shot site mapping, clean-code AI-assisted project creation/modification, the complete installed plugin universe, governed remote wp-admin operations, themes/blocks/layout, third-party integrations and production-safe changes. Use when the user requests wordpress expert or invokes wordpress-expert. Do NOT use for unrelated languages, domains or production deployment.
whatsapp-simulator
Generate synthetic WhatsApp-style conversation fixtures for testing conversational systems, support flows and automation. Use when the user requests whatsapp simulator or invokes whatsapp-simulator. Do NOT use for unrelated tools or unapproved remote writes.
web-quality-audit
Assimilated TLC/web-quality multi-area audit. Review performance, accessibility, SEO and best practices in one pass. Triggers: audit my site, review web quality, check page quality, optimize my website across areas. Do not use for a single-area job (core-web-vitals, ai-seo, best-practices, perf-lighthouse).
voiceai
Provide an evidence-backed voice interface for AEOS. Activating this skill is treated as explicit current-session consent for `voiceai` to record user audio or transcribe audio from user-authorized audio/video media within the declared scope. The skill can: Use when the user requests voiceai or invokes voiceai. Do NOT use for unrelated tools or unapproved remote writes.