@skillsmith/mcp-server 0.2.2 → 0.3.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +1 -1
- package/dist/.tsbuildinfo +1 -1
- package/dist/src/context.d.ts +23 -1
- package/dist/src/context.d.ts.map +1 -1
- package/dist/src/context.js +41 -1
- package/dist/src/context.js.map +1 -1
- package/dist/src/index.d.ts +1 -0
- package/dist/src/index.d.ts.map +1 -1
- package/dist/src/index.js +58 -0
- package/dist/src/index.js.map +1 -1
- package/dist/src/onboarding/first-run.d.ts +1 -0
- package/dist/src/onboarding/first-run.d.ts.map +1 -1
- package/dist/src/onboarding/first-run.js +2 -0
- package/dist/src/onboarding/first-run.js.map +1 -1
- package/dist/src/onboarding/install-assets.d.ts +27 -0
- package/dist/src/onboarding/install-assets.d.ts.map +1 -0
- package/dist/src/onboarding/install-assets.js +128 -0
- package/dist/src/onboarding/install-assets.js.map +1 -0
- package/dist/tests/onboarding/first-run.test.js +8 -2
- package/dist/tests/onboarding/first-run.test.js.map +1 -1
- package/package.json +3 -2
- package/src/assets/docs/USER_GUIDE.md +192 -0
- package/src/assets/skills/skillsmith/SKILL.md +147 -0
- package/src/assets/skills/skillsmith/docs/QUOTAS.md +182 -0
- package/src/assets/skills/skillsmith/docs/SECURITY.md +174 -0
- package/src/assets/skills/skillsmith/docs/TRUST_TIERS.md +142 -0
|
@@ -0,0 +1,142 @@
|
|
|
1
|
+
# Trust Tiers
|
|
2
|
+
|
|
3
|
+
Skillsmith uses a four-tier trust system to help you evaluate skill safety before installation.
|
|
4
|
+
|
|
5
|
+
## Tier Overview
|
|
6
|
+
|
|
7
|
+
| Tier | Badge | Auto-Install | Review Required |
|
|
8
|
+
|------|-------|--------------|-----------------|
|
|
9
|
+
| **Official** | Green checkmark | Yes | No |
|
|
10
|
+
| **Verified** | Blue checkmark | Yes | No |
|
|
11
|
+
| **Community** | Yellow circle | No | Recommended |
|
|
12
|
+
| **Unverified** | Red warning | No | Required |
|
|
13
|
+
|
|
14
|
+
## Official Tier
|
|
15
|
+
|
|
16
|
+
**What it means**: Published by Anthropic or trusted partners. Undergoes full security review.
|
|
17
|
+
|
|
18
|
+
**Requirements**:
|
|
19
|
+
- Published under `anthropic/` namespace
|
|
20
|
+
- Full code review by Anthropic security team
|
|
21
|
+
- Cryptographic signing (planned)
|
|
22
|
+
- Automatic updates deployed
|
|
23
|
+
|
|
24
|
+
**Examples**: `anthropic/varlock`, `anthropic/commit`, `anthropic/governance`
|
|
25
|
+
|
|
26
|
+
**When to install**: Always safe. These skills are maintained by Anthropic.
|
|
27
|
+
|
|
28
|
+
## Verified Tier
|
|
29
|
+
|
|
30
|
+
**What it means**: Publisher identity verified, skill meets quality and age requirements.
|
|
31
|
+
|
|
32
|
+
**Requirements**:
|
|
33
|
+
- Publisher identity verified via GitHub OAuth
|
|
34
|
+
- Automated security scan passed with no critical/high findings
|
|
35
|
+
- Minimum 10 GitHub stars
|
|
36
|
+
- Published for at least 30 days
|
|
37
|
+
- Has valid license file
|
|
38
|
+
- Complete README and SKILL.md
|
|
39
|
+
|
|
40
|
+
**Verification Process**:
|
|
41
|
+
1. Publisher submits verification request
|
|
42
|
+
2. Automated scan runs
|
|
43
|
+
3. Identity verification via GitHub
|
|
44
|
+
4. Manual review for edge cases
|
|
45
|
+
5. Verified badge granted (renewable annually)
|
|
46
|
+
|
|
47
|
+
**When to install**: Generally safe. Publisher is accountable for the skill.
|
|
48
|
+
|
|
49
|
+
## Community Tier
|
|
50
|
+
|
|
51
|
+
**What it means**: Passed basic security scan and has required metadata.
|
|
52
|
+
|
|
53
|
+
**Requirements**:
|
|
54
|
+
- Security scan passed (no critical findings)
|
|
55
|
+
- Valid SKILL.md with proper frontmatter
|
|
56
|
+
- Has LICENSE file
|
|
57
|
+
- Has README.md
|
|
58
|
+
- No blocklist matches
|
|
59
|
+
|
|
60
|
+
**What Community tier does NOT guarantee**:
|
|
61
|
+
- Publisher identity
|
|
62
|
+
- Code quality
|
|
63
|
+
- Ongoing maintenance
|
|
64
|
+
- No subtle security issues
|
|
65
|
+
|
|
66
|
+
**When to install**: Review skill content first. Check the author's GitHub profile and other projects.
|
|
67
|
+
|
|
68
|
+
## Unverified Tier
|
|
69
|
+
|
|
70
|
+
**What it means**: No verification performed. Could be newly published, failed scan, or intentionally unverified.
|
|
71
|
+
|
|
72
|
+
**Why a skill might be Unverified**:
|
|
73
|
+
- Just published (hasn't been scanned yet)
|
|
74
|
+
- Failed security scan
|
|
75
|
+
- Missing required files (LICENSE, README)
|
|
76
|
+
- Author hasn't submitted for verification
|
|
77
|
+
- Quarantined for suspicious activity
|
|
78
|
+
|
|
79
|
+
**When to install**: Only if you personally know and trust the author, or you've manually reviewed all code.
|
|
80
|
+
|
|
81
|
+
**Warning**: Unverified skills require explicit confirmation:
|
|
82
|
+
```
|
|
83
|
+
This skill is unverified. Are you sure you want to install? (y/N)
|
|
84
|
+
```
|
|
85
|
+
|
|
86
|
+
## Tier Transitions
|
|
87
|
+
|
|
88
|
+
### Upgrading from Community to Verified
|
|
89
|
+
1. Ensure skill meets all Verified requirements
|
|
90
|
+
2. Submit verification request at skillsmith.app/verify
|
|
91
|
+
3. Complete identity verification
|
|
92
|
+
4. Wait for review (typically 2-5 business days)
|
|
93
|
+
|
|
94
|
+
### Downgrades
|
|
95
|
+
Skills can be downgraded if:
|
|
96
|
+
- Security scan fails on update
|
|
97
|
+
- Publisher verification expires
|
|
98
|
+
- Reports of malicious behavior
|
|
99
|
+
- Author requests removal
|
|
100
|
+
|
|
101
|
+
## Filtering by Trust Tier
|
|
102
|
+
|
|
103
|
+
Use the `trust_tier` filter in searches:
|
|
104
|
+
|
|
105
|
+
```
|
|
106
|
+
"Find verified testing skills"
|
|
107
|
+
→ search(query="testing", trust_tier="verified")
|
|
108
|
+
|
|
109
|
+
"Show only official skills"
|
|
110
|
+
→ search(query="*", trust_tier="official")
|
|
111
|
+
```
|
|
112
|
+
|
|
113
|
+
## Trust Tier API
|
|
114
|
+
|
|
115
|
+
The `get_skill` tool returns trust information:
|
|
116
|
+
|
|
117
|
+
```json
|
|
118
|
+
{
|
|
119
|
+
"id": "community/jest-helper",
|
|
120
|
+
"trustTier": "verified",
|
|
121
|
+
"publisherVerified": true,
|
|
122
|
+
"scanPassed": true,
|
|
123
|
+
"scanDate": "2026-01-10",
|
|
124
|
+
"stars": 47,
|
|
125
|
+
"publishedDays": 89
|
|
126
|
+
}
|
|
127
|
+
```
|
|
128
|
+
|
|
129
|
+
## Recommendations by Use Case
|
|
130
|
+
|
|
131
|
+
| Scenario | Recommended Minimum Tier |
|
|
132
|
+
|----------|-------------------------|
|
|
133
|
+
| Production code | Verified or Official |
|
|
134
|
+
| Personal projects | Community or higher |
|
|
135
|
+
| Experimentation | Any (with review) |
|
|
136
|
+
| Enterprise/regulated | Official only |
|
|
137
|
+
|
|
138
|
+
## Questions?
|
|
139
|
+
|
|
140
|
+
- How do I get verified? See skillsmith.app/verify
|
|
141
|
+
- Report a suspicious skill: security@skillsmith.app
|
|
142
|
+
- Request tier review: support@skillsmith.app
|