cleanexpo/pi-dev-ops/agent-expert

Claude Code and agent skill by cleanexpo/pi-dev-ops

Trust tier 3
  • community
Source check pending

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

Act-Learn-Reuse cycle for agent improvement over time. Use when a task finishes and its lessons should be captured and reused, or when a user asks to log lessons to .harness/lessons.jsonl, inject the top relevant lessons into the next task, or record that a skill's own SKILL.md was unclear or caused a mistake (via the applies_to_skill field).

Install

npx vskill@latest install cleanexpo/pi-dev-ops/agent-expert

SKILL.md

The same SKILL.md was published earlier as cleanexpo/pi-dev-ops/agent-expert.

Agent Experts

The Cycle

  1. ACT - Execute the task
  2. LEARN - Extract lessons (patterns, pitfalls, context, tools, conventions)
  3. REUSE - Inject relevant lessons into next task

Store lessons in .harness/lessons.jsonl. Inject top 5 most relevant per task.

Skill-process lessons

If the lesson is about a skill's own process or instructions — its SKILL.md was unclear, missing a step, contradicted itself, or its guidance is what caused the mistake — rather than about the task's domain content, add one extra field: applies_to_skill, set to that skill's exact folder name under skills/ (e.g. "agent-expert", "skill-authoring-standard"). Omit this field entirely for ordinary domain lessons (a code bug, an infra misconfiguration, a business-logic error) — it is optional and skill-scoped lessons are the minority case.

Example — a domain lesson (no applies_to_skill):

json
{"ts": "2026-07-03T02:00:00Z", "source": "architecture-review", "category": "security", "lesson": "Webhook signature verification must use hmac.compare_digest, not ==.", "severity": "warn"}

Example — a skill-process lesson (applies_to_skill set):

json
{"ts": "2026-07-03T02:00:00Z", "source": "linear-task-processor", "category": "process", "applies_to_skill": "skill-authoring-standard", "lesson": "The review-checklist.md line-count cap (200 lines) doesn't say whether YAML frontmatter counts toward the limit — two agents interpreted this differently in the same week. Clarify in the checklist itself.", "severity": "warn"}

scripts/scan_skill_lessons.py reads this field to detect when a skill's own instructions need a fix — see that script for the threshold and PR flow.

Security

  • Security scanPassed, 90 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/cleanexpo/pi-dev-ops/agent-expert/badge)](https://verified-skill.com/skills/cleanexpo/pi-dev-ops/agent-expert)

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