Agent skill · Audited from source

search-skill

skill · gbsoss/skill-from-mastersMIT licenceAudited 20 Oct 2026 · source 0144be4307b8

★ 1,591pushed 246 days ago
✓SAFEGrade B · 89/100 · 20 Oct 2026 · 0144be4
1,591repo stars · whole gbsoss/skill-from-masters

Is it safe? Yes — SAFE, grade B (89/100). We read this skill's source at commit 0144be4307b8 and found no filesystem, network or shell access and no hard-coded secrets. What it does: Search and recommend Claude Code skills from trusted marketplaces

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “search-skill (gbsoss/skill-from-masters)”. Retrieved 20 Oct 2026. https://modelsatlas.com/skills/gbsoss/search-skill

3 skills are called “search-skill”. You are viewing gbsoss/skill-from-masters.

Trust audit

Audit 0.4.1 · source read 20 Oct 2026

Trust score89/100✓SAFE B
0 critical0 high0 medium0 low0 info

SAFE means nothing in the source contradicts what the skill says it does. Grade B is shared by 13,985 of 14,816 audited skills.

What the source does

  • FilesystemNone observedNo reads or writes in the source
  • NetworkNone observedNo outbound calls in the source
  • ShellNone observedNo commands run in the source
  • DependenciesPinnedNothing floats
  • SecretsNone foundNo keys in source

Five audit layers

LayerWhat we checkResultFindings
L0 ProvenanceRepo owner, licence, archive status, activityPASS0
L1 Static sourceSecrets, path traversal, supply chain, hidden codeNA0
L2 InstructionsPrompt injection, authority framing, hidden directivesPASS0
L3 InventoryThe files and declarations match the sourcePASS0
L4 SandboxRun in isolation and watch behaviourSKIPPED why0

Gates applied: no behavioural pass.

Findings (0)

No findings at any severity. We re-audit when the source changes.

Audit 0.4.1 · source sha 0144be4307b8 · 20 Oct 2026Report an issue or request a re-scan

Provenance

Who published it, and what the verdict is pinned to

  • Publishergbsoss GitHub organizationOwner of the repository this skill is read from
  • Source foldergbsoss/skill-from-masters/skills/search-skillThe exact folder we audited
  • LicenceMITAs the repository states it
  • Commit pinned for this verdict0144be4307b8A new audit runs when the source changes
  • Repository statusNot archivedLast push 246 days ago
  • Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

What it tells the agent

Full source on GitHub ↗

Its description: Search and recommend Claude Code skills from trusted marketplaces

---
name: search-skill
description: Search and recommend Claude Code skills from trusted marketplaces
model: sonnet
---

# Search Skill

Search and recommend Claude Code Skills from trusted marketplaces based on user requirements.

## When to Use

When users describe a need and want to find an existing Skill to solve it.

Examples:
- "Is there a skill that can auto-generate changelogs?"
- "Find me a skill for frontend design"
- "I need a skill that can automate browser actions"

## Data Sources (by trust level)

### Tier 1 - Official / High Trust (show first)
| Source | URL | Notes |
|--------|-----|-------|
| anthropics/skills | github.com/anthropics/skills | Official examples, most reliable |
| ComposioHQ/awesome-claude-skills | github.com/ComposioHQ/awesome-claude-skills | Hand-picked, 12k+ stars |

### Tier 2 - Community Curated (secondary)
| Source | URL | Notes |
|--------|-----|-------|
| travisvn/awesome-claude-skills | github.com/travisvn/awesome-claude-skills | Community curated, 21k+ stars |
| skills.sh | skills.sh | Vercel's official directory |

### Tier 2.5 - Large Community Registry
| Source | URL | Notes |
|--------|-----|-------|
| ClawHub | clawhub.ai | OpenClaw official skill registry, 5000+ community skills, vector search powered |

### Tier 3 - Aggregators (use with caution)
| Source | URL | Notes |
|--------|-----|-------|
| skillsmp.com | skillsmp.com | Auto-scraped, requires extra filtering |

## Search Process

### Step 1: Parse User Intent

Extract from user description:
- Core functionality keywords (e.g., changelog, browser, frontend)
- Use case (e.g., development, testing, design)
- Special requirements (e.g., language support, specific framework)

### Step 2: Multi-Source Search

**IMPORTANT: Only search these 6 sources. Do NOT search the entire internet.**

Search by priority:

```
1. Search Tier 1 (official/high trust) first
2. If fewer than 5 results, continue to Tier 2
3. If still insufficient, search Tier 2.5 (ClawHub - large volume, check quality)
4. If still insufficient, search Tier 3 with strict filtering
5. If still nothing found, tell user honestly - do NOT expand to other sources
```

Allowed search queries (use `site:` to restrict):
```
site:github.com/anthropics/skills {keywords}
site:github.com/ComposioHQ/awesome-claude-skills {keywords}
site:github.com/travisvn/awesome-claude-skills {keywords}
site:skills.sh {keywords}
site:clawhub.ai {keywords}
site:skillsmp.com {keywords}
```

Search methods:
- GitHub repos: Use `site:github.com/{repo}` to restrict search scope
- skills.sh: WebFetch to scrape search results from skills.sh only
- ClawHub: WebFetch `clawhub.ai/skills?q={keywords}` to search the registry
- skillsmp.com: WebFetch with additional verification

**Do NOT:**
- Search the entire web
- Use broad queries without `site:` restriction
- Include results from unknown sources

### Step 3: Quality Filtering (Critical)

**Must filter out the following:**

| Filter Condition | Reason |
|------------------|--------|
| GitHub stars < 10 | Not community verified |
| Last update > 6 months ago | Possibly abandoned |
| No SKILL.md file | Non-standard format |
| README too sparse | Quality concerns |
| Contains suspicious code patterns | Security risk |

**Security checks:**
- Requests sensitive permissions (e.g., ~/.ssh, env variables)
- External network requests to unknown domains
- Contains eval() or dynamic code execution
- Modifies system files

### Step 4: Rank Results

Scoring formula:
```
Score = Source Weight × 0.4 + Stars Weight × 0.3 + Recency Weight × 0.2 + Relevance × 0.1

Source weights:
- Tier 1: 1.0
- Tier 2: 0.7
- Tier 2.5 (ClawHub): 0.55
- Tier 3: 0.4
```

### Step 5: Format Output

Return Top 5-10 results:

```markdown
## Found X relevant Skills

### Recommended
1. **[skill-name](github-url)** - Source: anthropics/skills
   - Function: xxx
   - Stars: xxx | Last updated: xxx
   - Install: `/plugin marketplace add xxx`

### Worth considering
2. **[skill-name](github-url)** - Source: ComposioHQ
   ...

### Not recommended (for reference only)
- [skill-name](url) - Reason: low stars / not maintained
```

## Example

**User**: Is there a skill that helps write commit messages?

**Search process**:
1. Extract keywords: commit, message, git
2. Search Tier 1: Found git-commit-assistant in anthropics/skills
3. Search Tier 2: Found semantic-commit in ComposioHQ
4. Filter: Exclude results with stars < 10
5. Rank: Official sources first

**Output**:
```
## Found 3 relevant Skills

### Recommended
1. **git-commit-assistant** - Source: anthropics/skills (official)
   - Function: Generate semantic commit messages
   - Install: `/plugin marketplace add anthropics/claude-code`

2. **semantic-commit** - Source: ComposioHQ
   - Function: Follow conventional commits spec
   - Stars: 890 | Last updated: 2 weeks ago
```

## Important Notes

1. **Never recommend unverified Skills** - Better to recommend fewer than to recommend risky ones
2. **Stay cautious with Tier 3 sources** - Results from skillsmp.com must be double-checked
3. **If nothing suitable is found** - Tell the user honestly, suggest using skill-from-masters to create their own
4. **Security concerns** - Clearly inform users of risks, let them decide

Excerpt, first 5,246 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

Install

Pinned to the audited commit 0144be4

Download folder ↗
git clone https://github.com/gbsoss/skill-from-masters
git -C skill-from-masters checkout 0144be4307b8

The author documents no install command, so this uses the repository directly. Our verdict covers commit 0144be4307b8 only.

Works in which agents?

What its own text says, not a test

AgentStatusMore
Claude CodeMentioned in its text, not testedSkills for Claude Code · About Claude Code
OpenClawMentioned in its text, not testedSkills for OpenClaw · About OpenClaw
“Mentioned” means the skill’s text names the agent. We have not run it there.

Audit history

Every audit, every source change

  1. ✓SAFE B 89 / 100

    first audit · source 0144be4307b8

README badge

For maintainers · shows the latest audit

ModelsAtlas: SAFE BMODELSATLASSAFE · B
[![ModelsAtlas: SAFE B](https://modelsatlas.com/skills/gbsoss/search-skill/badge.svg)](https://modelsatlas.com/skills/gbsoss/search-skill/)

The badge shows our verdict and grade for this skill, and links back to this page. It changes when the audit does.

Similar skills

Add this badge to your README

[![ModelsAtlas: SAFE B](https://modelsatlas.com/skills/gbsoss/search-skill/badge.svg)](https://modelsatlas.com/skills/gbsoss/search-skill/)

Questions about search-skill

What does the search-skill skill do?
Stand on the shoulders of giants — Create AI skills built on proven methodologies from domain experts.
Is search-skill safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can search-skill access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
Which assistants does search-skill work with?
Its documentation mentions claude-code and openclaw. That is what the text claims, not a compatibility test we ran.
How current is this page?
The grade is for one exact copy of the source (0144be4307b8), read on 2026-10-20. The repository is watched, and a new audit runs when it changes — this is the first audit.

Sources

Spotted something wrong? Report a correction · Methodology

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.