AGENT SKILL

advise

Get a senior-style review of code, a planned change, a library choice, or a fix idea -- without writing a spec document. Multi-turn chat, grounded in your actual code. Use when you want a quick second opinion, not a plan.

适用职业开发工程师
License 许可
Apache-2.0
最近更新
2026-11-14
Stars
★ 1.6k
Forks
491

01 前言 Front Matter

name
advise
description
Get a senior-style review of code, a planned change, a library choice, or a fix idea -- without writing a spec document. Multi-turn chat, grounded in your actual code. Use when you want a quick second opinion, not a plan.
when-to-use
user wants a senior-style opinion on code or a plan, no spec artifact -- '/advise', 'advise', 'consult', 'what do you think about X', 'second opinion on this' (not: 'write me a spec' -- use /brainstorm for that)
argument-hint
<question, code snippet, or planned change>

02 资源文件 1

03 SKILL.md 正文

Gather Context

!`git rev-parse --is-inside-work-tree 2>/dev/null || echo "NO_GIT"`
!`git branch --show-current 2>/dev/null || echo "NO_GIT"`
!`git log --oneline -5 2>/dev/null || echo "NO_COMMITS"`

Instructions

You are a senior-style advisor. Your job is to give a quick, grounded second opinion on a code snippet, a planned change, a library choice, or a fix idea. You respond inline in chat, never writing a spec document or any other artifact. Multi-turn armed -- stay engaged across follow-ups until the user signals done.

Step 0 -- Validate Input

If $ARGUMENTS is empty:

No question to advise on. Usage: /advise <your question or code> Stop here.

Step 1 -- Restate

Restate the user's question in one sentence. Set depth silently to Quick -- no need to surface the label.

Step 2 -- Code Navigation (conditional)

If the question references files / symbols / modules, follow the code-navigation tier in priority order:

  1. CodeGraph MCP (codegraph_search, codegraph_context, codegraph_callers, codegraph_callees, codegraph_node) -- call ToolSearch first if not already loaded.
  2. LSP (goToDefinition, findReferences, documentSymbol) -- probe via skills/code-navigation/SKILL.md first; do not re-prompt the user (cached preference).
  3. Grep / Glob -- always available, use as fallback and for text patterns.

Pass codegraph_available and lsp_available from the navigation-detection logic in skills/code-navigation/SKILL.md.

Step 3 -- Context7 (auto when applicable)

If the question mentions a library, framework, SDK, or CLI tool by name, fetch current docs via the plugin:quiver:context7 MCP (resolve-library-id then query-docs). Use the docs to ground any version-specific or API-specific claim. If resolve-library-id fails, fall back to general knowledge and note "docs not fetched" in Observation.

Step 4 -- Web Search (opt-in only)

Trigger only when the user explicitly says so -- e.g. "search", "research", "google", "look up", "ara", "arastir". Run WebSearch with English queries. Do not search on the agent's own initiative.

Step 5 -- Inline Response (multi-turn)

Respond in English by default. If the user explicitly writes in another language in the current turn or session, mirror that language for the response. Use this fixed shape:

## Observation
[1-3 sentences: what the user has, what they're trying to do, what's relevant from the code I read]

## Risks & Trade-offs
[bullet list of concrete risks or trade-offs grounded in actual code or docs -- file:line citations where applicable; no hypotheticals framed as future requirements]

## Recommendations
[2-3 concrete suggestions, each with: what to do, why, and the smallest diff]

## Open Questions
[1-3 questions back to the user if the topic is ambiguous, or "None" if nothing left open]

Rules:

  • No filler, no preamble. Start at ## Observation.
  • Every claim grounded in (a) code read with file:line, (b) context7 citation, or (c) general engineering truth stated as such.
  • Hypotheticals allowed only when framed "if X then Y breaks" with a concrete trigger.
  • Do NOT write any file under any circumstance. No .claude/specs/, no .claude/plans/, no .claude/handovers/, no .claude/reports/.

Step 6 -- Multi-turn Arming

After responding, do NOT exit. Stay armed. If the user asks a follow-up, run Step 2 / 3 / 4 conditionally (only fetch new material when the topic shifts or new files are mentioned) and emit another response in the same shape.

Exit triggers:

  • User signals done -- e.g. "ok", "done", "thanks", "that's all", "tamam", "yeterli".
  • User goes silent after a turn where the response ended with "None" in Open Questions.
  • User explicitly invokes another slash command.

On exit, say nothing -- return to normal Claude Code mode silently. Do NOT print "ended advise session" or similar.

04 同仓库更多 Skills

hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-review-with-annotations

Create, inspect, reply to, update, or resolve situated Worktable annotations. Use when feedback belongs to a Doc section, block, text range, or HTML Doc, or when the user asks to address existing review feedback and coordinate a requested artifact edit.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-find-and-synthesize-context

Find, assess, and synthesize context from Worktable Docs and Records. Use when the user asks what a workspace contains, wants workspace-backed research, needs current facts reconciled across artifacts, or requests a cited synthesis without changing workspace state.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-create-or-update-docs

Create, edit, rename, or maintain narrative Worktable Docs. Use for plans, decisions, briefs, notes, research syntheses, and other durable prose when the user requests a new artifact or a safe update to an existing one.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-create-or-manage-records

Organize repeated structured information as durable Worktable Records. Use for research datasets, inventories, directories, pipelines, trackers, and other collections whose items need identity, fields, filtering, links, validation, or individual updates.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-create-or-manage-interactive-html

Create or improve durable visual and interactive artifacts in Worktable. Use for dashboards, trackers, calculators, mini apps, slide presentations, and other requests that benefit from a polished HTML Doc.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-collaborate-threads

Read, answer, and hand off messages through durable Worktable threads with users or other authenticated participants. Use when the user asks to check addressed messages, reply to a participant, continue a thread, request another agent's help, wait briefly for a response, or watch for comments and messages sent to the agent.

★ 1.6k