bish-x/bx-dev-skill/skill-finder

Claude Code and agent skill by bish-x/bx-dev-skill

Trust tier 3
  • community
Source check pending

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

Skill library router. Searches the skill library by task context and loads matching skills on demand. Always active.

Install

npx vskill@latest install bish-x/bx-dev-skill/skill-finder

SKILL.md

Skill Finder

You have access to a skill library in this repository's skill-library/ folder, or the equivalent installed skill-library root in your Codex setup. The root INDEX.md is the single source of truth for categories and bundled skills.

When to Search

Before starting any substantive task (coding, writing, reviewing, designing, planning, debugging, marketing), search for a matching skill.

Skip search for trivial tasks:

  • Single-line fixes with explicit location ("fix typo on line 42")
  • Simple read/info requests ("show me file X", "what does this function do?")
  • One-off shell commands ("run tests", "git status")

NOT trivial (do search): "edit README", "update docs", "refactor X" — scope is unclear, skill may help.

How to Search

Step 1 — Pick category

Read <skill-library-root>/INDEX.md and match the task to a category:

CategoryTriggers
architectureDDD, bounded contexts, domain modeling, architecture boundaries
backendAPI, database, Postgres, Supabase, SQL, server, security, DevOps, Docker, K8s, CI/CD, debugging, monitoring, MCP
frontendReact, Next.js, UI, CSS, Tailwind, accessibility, design, mobile, React Native, Remotion, video
review-qacode review, website audit, user research, QA, performance audit, E2E testing, Playwright
marketingcopy, SEO, CRO, ads, email, pricing, launch, content strategy, analytics, internal comms
generalplanning, brainstorming, bash, git, diagrams, video/FFmpeg, legal, product work
systemCodex/OpenAI docs, image generation, plugins, skill creation/installation, migration, routing

If no category matches — skip, no skill needed.

Step 2 — Find skill

Use the "Full Index" section in <skill-library-root>/INDEX.md. Each entry is listed as:

`category/skill-name`

Match by category, skill name, and task context. If still ambiguous, open the smallest plausible SKILL.md and check its frontmatter description before loading more.

Step 3 — Load skill

Read <skill-library-root>/<category>/<skill>/SKILL.md and follow its instructions.

Budget rule: total loaded SKILL.md content must not exceed 1500 lines per task. If a skill is large, read only the relevant section after checking its frontmatter and headings.

Disambiguation

If prompt is too generic and matches multiple skills, resolve by context:

Generic termClarify by context
"review"git diff → code-review-expert; codebase quality → code-quality-review; live URL → audit-website
"performance"React/bundle → vercel-react-best-practices; Lighthouse/CWV → performance-audit; monitoring/SLO → observability-design
"security"source review → code-review-expert; live website pentest → security-pentest; payments/PCI → pci-compliance; secrets/.env → secrets-vault-manager
"optimize form"signup/registration → signup-flow-cro; lead capture/contact → form-cro
"email"drip sequence → email-sequence; single copy → copywriting; analytics → analytics-tracking
"document"write RFC/spec → engineering/writing-plans or general/spec-driven-workflow; legal file review → general/legal-review; meeting transcript analysis → general/meeting-analyzer; otherwise use the built-in file tools available in the current environment

If still ambiguous after this table — ask the user one short clarifying question before loading.

Multi-Category Tasks

If task spans 2+ categories, load skills in this order:

  1. Plan — general (brainstorming, consultancy-practices)
  2. Context — marketing/product-marketing-context, general/product-discovery, engineering/context-engineering
  3. Build — backend, frontend, system/general support as relevant
  4. Verify — review-qa, engineering/systematic-debugging
  5. Ship — marketing (launch, copy, SEO)

Within one phase, load the most specific skill first.

Budget Management

  • Max budget: 1500 lines of SKILL.md content per task
  • If no skill matches — proceed without one, don't force it
  • Never load a skill "just in case" — only when there's a clear match
  • If task spans multiple categories — load from each, respecting the budget
  • If compound task exceeds 1500 lines: load skills sequentially by phase — execute phase 1, then unload and load phase 2. Never load all at once.

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
0

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/bish-x/bx-dev-skill/skill-finder/badge)](https://verified-skill.com/skills/bish-x/bx-dev-skill/skill-finder)

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