菜单
Approve pending library requests and discovered tools. Updates manifest.json with approval metadata. Handles both library approvals and Tier 2b tool acceptance from discover().
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
| name | raven-approve |
| description | Approve pending library requests and discovered tools. Updates manifest.json with approval metadata. Handles both library approvals and Tier 2b tool acceptance from discover(). |
Usage: /approve {library} {version}
Steps:
Dev must pull latest manifest before proceeding.
Usage: /approve tool {tool_name} or user says "approve" after discover() presents a tool.
Steps:
.raven/.cache/discovery/{tool-slug}.jsonstack.tools:
{
"name": "{tool_name}",
"source": "{mcp-registry / github / vendor}",
"trust": "{HIGH/MEDIUM/LOW}",
"approved_by": "{user}",
"approved_at": "{ISO8601}",
"discovered_at": "{ISO8601}",
"review_by": "{approved_at + 90 days}"
}
When a session starts and manifest has tools with review_by < today:
⚠️ TOOL REVIEW DUE:
{tool_name} — approved {date}, review due {date}
Trust: {level} | Source: {source}
→ Say "renew" to extend 90 days, "remove" to revoke, "info" to check current status.
Usage: /approve remove {tool_name}
Steps:
Initialize Raven for a new project. Andie-conversational — auto-scans project, asks at most 2 questions, generates manifest.json, validates, and commits with audit trail.
USE PROACTIVELY whenever the user asks for: planning, design, architecture decision, tradeoff analysis, comparing approaches, strategy, system design, refactor scope, deciding what to build, or any non-trivial request needing clarification. Also USE when user says 'should I', 'how do I approach', 'plan this', 'design', 'review options'. Compact plan-first orchestration. Routes work, runs triad (Functional/Technical/Data), HITL gated, OODA loop. Hands off plans, never implements. Brownfield bugs → andie-jr.
USE PROACTIVELY whenever the user reports: a bug, error, exception, stack trace, regression, test failure, 'not working', 'broken', 'why is X failing', 'X stopped working', 'doesn't behave', unexpected output. Brownfield debug assistant. Max 2 rounds, max 3 roles, returns problem · root cause · fix · why · audit note · commit suggestion. Do NOT use for greenfield design — that goes to andie.
USE PROACTIVELY whenever the user asks for: planning, design, architecture decision, tradeoff analysis, comparing approaches, strategy, system design, refactor scope, deciding what to build, or any non-trivial request needing clarification. Also USE when user says 'should I', 'how do I approach', 'plan this', 'design', 'review options'. Compact plan-first orchestration. Routes work, runs triad (Functional/Technical/Data), HITL gated, OODA loop. Hands off plans, never implements. Brownfield bugs → andie-jr.
Tool and product registry for Andie pre-flight. Maps problem types to tool options with tradeoffs. Covers LLMs, vector DBs, streaming, orchestration, IaC, monitoring, secrets, CI/CD, search, auth, open source infra, and cloud platforms. Load when a tool selection decision is in scope.
Tool and product registry for Andie pre-flight. Maps problem types to tool options with tradeoffs. Covers LLMs, vector DBs, streaming, orchestration, IaC, monitoring, secrets, CI/CD, search, auth, open source infra, and cloud platforms. Load when a tool selection decision is in scope.