AgentStack
SKILL verified MIT Self-run

Skill Navigator

skill-fxyadela-skill-navigator-skill-navigator · by fxyadela

Use this skill when the user is unsure which local Codex skill to use, asks for skill recommendations, wants a guided route through available skills, or wants to turn a vague task into the right skill invocation. It maps the user's task to one or more installed skills and gives a copy-ready next prompt.

No reviews yet
0 installs
4 views
0.0% view→install

Install

$ agentstack add skill-fxyadela-skill-navigator-skill-navigator

✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

Are you the author of Skill Navigator? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Skill Navigator

This is a routing skill. It should reduce the user's need to remember skill names.

Core Rule

Do not dump the full skill list by default. First classify the user's intent, then recommend the smallest useful set of skills.

Workflow

  1. Restate the user's real task in one sentence.
  2. Pick one primary skill. Add a second skill only if the workflow naturally needs chaining.
  3. Explain the choice briefly: what this skill does that the others do not.
  4. Give a copy-ready prompt using the $skill-name form.
  5. If the task is under-specified, ask at most two questions that materially affect the route.

Routing Map

  • Source/content extraction from a URL: $link-extractor
  • YouTube transcript or subtitles: $youtube-transcript
  • Claim/source/evidence verification: $claim-forensics
  • Question-first reading and cognitive navigation: $sinan
  • Low-cognitive-load learning notes: $knowledge-juicer
  • Chinese spoken scripts or creator-style rewrites: $koubo-script-writer
  • Video material generation prompts from scripts: $storyboard-prompt-writer
  • Real video material sourcing and editing plans: $koubo-video-material-finder
  • Short-video covers, thumbnails, poster prompts: $cover-prompt-director
  • Static visual art, posters, PNG/PDF design: $canvas-design
  • Remotion videos with transparent overlays: $remotion-overlay-export
  • Contract or vendor-form filling from local private profile: $contract-assistant
  • IMA notes, knowledge bases, uploads, and searches: $ima-skill
  • Paragraph cleanup and hard-line-break repair: $paragraph-reflow
  • ArXiv paper search: $arxiv-watcher
  • GitHub AI project trends: $github-ai-trends
  • Blog and RSS monitoring: $blogwatcher
  • AI builder digest from X or YouTube podcasts: $follow-builders

Output Shape

**我会用**
`$skill-name`

**为什么**
...

**下一步直接这样说**
Use $skill-name to ...

**如果要串联**
先用 `$skill-a` ...,再用 `$skill-b` ...

Judgment Rules

  • If the user says "找来源", "出处", "这是谁说的", or "有没有证据", prefer $claim-forensics over generic search.
  • If the user gives a URL but asks only to read it, start with $link-extractor; if they ask whether claims are true, chain into $claim-forensics.
  • If the user asks for learning, distinguish $sinan from $knowledge-juicer: use $sinan when the user wants thinking navigation, and $knowledge-juicer when they want a lightweight study artifact.
  • If the user asks for video work, distinguish writing, generation prompts, real素材 planning, and transparent animation before choosing.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet — be the first.

Versions

  • v0.1.0 Imported from the upstream source.