jleechanorg/jleechan-skills/advice

Claude Code and agent skill by jleechanorg/jleechan-skills

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Obtain independent, revision-bound second opinions with explicit coverage and a fail-closed approval quorum.

Install

npx vskill@latest install jleechanorg/jleechan-skills/advice

SKILL.md

Advice

  1. Ground the decision, constraints, exact revision or document hashes, complete changed files/diff and relevant evidence. Give reviewers pointers to readable artifacts, not the entire conversation or a framed question; the orchestrator answers the user's question in its synthesis.
  2. Select eligible reviewers through the canonical advice tiers before using parallelize-to-ceiling for any admitted native reviewer lanes within current authorized capacity. Reviewers must read the complete assigned material independently. Preserve their identity, revision, coverage and raw verdict as review evidence.
  3. Follow the canonical lifecycle-selected quorum and advice reviewer-class caps. One full-scope reviewer suffices only for an eligible one-vote route; stricter and mandatory two-vote obligations remain. Research and the orchestrator do not vote. A competent objection can block; unavailable or partial coverage cannot satisfy quorum. Report APPROVED, NOT APPROVED or WITHHELD, with findings and missing coverage.
  4. Run research only when a separate unanswered source question warrants a lane; keep it outside the approval quorum. External browser review is a separate web-advice route requiring the applicable recipient/content authorization and the canonical last-tier admission; availability alone grants no disclosure permission.
  5. Bind findings to paths/lines and the exact source revision. After a head move, inspect the delta and changed external artifacts. Material changes need fresh review. A nonbehavioral delta may be explicitly reaffirmed only with prior coverage, provenance and the new delta reviewed; never relabel old evidence.
  6. Preserve completed results and retry only concretely failed lanes. Read supporting artifacts before relying on a reviewer summary. Resolve findings or report why approval is withheld.
  7. Save the bounded synthesis in the authorized workspace. Publishing a gist, PR comment, or private source to an outside provider is not implicit. Missing transport means missing coverage, not permission to bypass access controls.
  8. Every reviewer gets a cold, verbatim, static-review-owned prompt (only its specified fields filled) plus a fixed per-channel preamble; the receipt records the exact prompt, and any deviation voids the vote. Load ${CLAUDE_HOME:-$HOME/.claude}/skills/advice/SKILL.md, the canonical owner, not this portable summary. Follow its §2 for capability tiers, at-most-one-Dot counting, aggregate static/Hermes caps and web-last fallback. This package does not ship that owner: if it is absent or resolves back to this summary, record scheduling UNAVAILABLE and the review WITHHELD; never invent the missing policy. Retain eligible earlier-tier votes under the canonical quorum, independence and substitution rules. Record each unavailable seat as host, CLI or static reviewer, attempted or skipped, time and reason, and tell the user. Static-review votes need the canonical /static-review skill's verbatim templates and preambles; this portable package does not ship them, so where that skill is not installed, record static review as UNAVAILABLE rather than improvising a prompt.

Security

  • Security scanPassed, 95 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
5

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/jleechanorg/jleechan-skills/advice/badge)](https://verified-skill.com/skills/jleechanorg/jleechan-skills/advice)

Something wrong with this skill or its page? Tell us.