agent-mart— agent skill

A browsable directory of open source Claude Code plugins

by brsbl·Agent Skill·github.com/brsbl/agent-mart

Is agent-mart safe to install?

SaferSkills independently audited agent-mart (Agent Skill) and scored it 45/100 (orange). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 8 high-severity and 1 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.

Score
45/100
●●●●●○○○○○
↑ +0 since first scan (45 → 45)Re-scan~30s
Latest scan
ScannedJun 28, 2026 · 30d ago
Scans run1 over 90 days
Detectors55 checks · 5 categories
Findings1 warnings · 8 high
EngineSaferSkills 2b638c6
View methodology →
SaferSkills installs
This week0
This month0
All time0
CategoryWeightCategory scoreContribution
Securityprompt, exec, net, exfil, eval
35%
0
0.0 pts
Supply chainhash, typosquat, maintainer, lockfile
20%
100
20.0 pts
Maintenancestaleness, pinning, CI
15%
100
15.0 pts
TransparencySKILL.md, perms, README
15%
100
15.0 pts
Communityinstalls, verify, response
15%
100
15.0 pts

Findings & checks · 9 flagged

Securityscore 0 · 9 findings
HIGHLong base64-encoded blob hidden in the skill documentationSS-SKILL-INJECT-B64-PAYLOAD-01 · Prompt injection · web/public/data/authors/ai-riksarkivet.json×2
HIGHonce decoded by the agent, an encoded payload has the same impact class as plain-text injection.
Why it matters

A base64 string of 128+ characters appears in a documentation file. Encoded prompt injection hides the hostile instruction in base64 — invisible to keyword filters — and relies on the agent's ability to decode it at runtime. There is no normal authoring reason to embed a multi-hundred-byte base64 blob in skill docs.

The exact value spotted
excerptweb/public/data/authors/ai-riksarkivet.json· json
20"files": {
21".claude-plugin/marketplace.json": "{\n \"name\": \"ra-mcp-plugins\",\n \"owner\": {\n
… (100 chars elided on L21)
22"README.md": "<div align=\"center\">\r\n <img src=\"https://raw.githubusercontent.com/AI-Ri
… (100 chars elided on L22)
23"plugins/ra-mcp-tools/README.md": "# Riksarkivet Tools Plugin for Claude Code\n\nA Claude Co
… (100 chars elided on L23)
24},
Occurrences
2 occurrences · first at L22, also L22
Show all 2 locations
Line
File
L22
web/public/data/authors/ai-riksarkivet.json
L22
web/public/data/authors/ai-riksarkivet.json
How to fix
Remove the encoded blob, or decode it and review what it actually contains.
  1. Decode the base64 string and confirm it is not an instruction directed at the agent.
  2. Move any legitimate binary or signature data into a dedicated file (*.sig, SIGNATURES) outside the documentation.
Framework references
OWASPLLM01ATLASAML.T0051
Trace & refs
ruleSS-SKILL-INJECT-B64-PAYLOAD-01sha25633481c138a8d7911rubric 365aacaView on GitHub
HIGHLong base64-encoded blob hidden in the skill documentationSS-SKILL-INJECT-B64-PAYLOAD-01 · Prompt injection · web/public/data/authors/artifex1.json
HIGHonce decoded by the agent, an encoded payload has the same impact class as plain-text injection.
Why it matters

A base64 string of 128+ characters appears in a documentation file. Encoded prompt injection hides the hostile instruction in base64 — invisible to keyword filters — and relies on the agent's ability to decode it at runtime. There is no normal authoring reason to embed a multi-hundred-byte base64 blob in skill docs.

The exact value spotted
excerptweb/public/data/authors/artifex1.json· json
21".claude-plugin/marketplace.json": "{\n \"name\": \"auditor-addon-marketplace\",\n \"owner
… (100 chars elided on L21)
22".claude-plugin/plugin.json": "{\n \"name\": \"auditor-addon\",\n \"version\": \"1.0.11\",
… (100 chars elided on L22)
23"README.md": "<div align=\"center\">\n\n<img src=\"./logo.svg\" alt=\"Auditor Addon Logo\" w
… (100 chars elided on L23)
24},
25"plugins": [
Occurrences
1 occurrence · at L23
How to fix
Remove the encoded blob, or decode it and review what it actually contains.
  1. Decode the base64 string and confirm it is not an instruction directed at the agent.
  2. Move any legitimate binary or signature data into a dedicated file (*.sig, SIGNATURES) outside the documentation.
Framework references
OWASPLLM01ATLASAML.T0051
Trace & refs
ruleSS-SKILL-INJECT-B64-PAYLOAD-01sha25617cccf92fae171c5rubric 365aacaView on GitHub
HIGHLong base64-encoded blob hidden in the skill documentationSS-SKILL-INJECT-B64-PAYLOAD-01 · Prompt injection · web/public/data/authors/earthmanweb.json
HIGHonce decoded by the agent, an encoded payload has the same impact class as plain-text injection.
Why it matters

A base64 string of 128+ characters appears in a documentation file. Encoded prompt injection hides the hostile instruction in base64 — invisible to keyword filters — and relies on the agent's ability to decode it at runtime. There is no normal authoring reason to embed a multi-hundred-byte base64 blob in skill docs.

The exact value spotted
excerptweb/public/data/authors/earthmanweb.json· json
66".claude-plugin/marketplace.json": "{\n \"name\": \"claude-flow-marketplace\",\n \"owner\"
… (100 chars elided on L66)
67".claude-plugin/plugin.json": "{\n \"name\": \"claude-flow\",\n \"version\": \"2.5.0\",\n
… (100 chars elided on L67)
68"README.md": "# 🌊 Claude-Flow v3: Enterprise AI Orchestration Platform\n\n<div align=\"cent
… (101 chars elided on L68)
69},
70"plugins": [
Occurrences
1 occurrence · at L68
How to fix
Remove the encoded blob, or decode it and review what it actually contains.
  1. Decode the base64 string and confirm it is not an instruction directed at the agent.
  2. Move any legitimate binary or signature data into a dedicated file (*.sig, SIGNATURES) outside the documentation.
Framework references
OWASPLLM01ATLASAML.T0051
Trace & refs
ruleSS-SKILL-INJECT-B64-PAYLOAD-01sha256a9a529cb9f0f7c0frubric 365aacaView on GitHub
HIGHLong base64-encoded blob hidden in the skill documentationSS-SKILL-INJECT-B64-PAYLOAD-01 · Prompt injection · web/public/data/authors/enessari.json
HIGHonce decoded by the agent, an encoded payload has the same impact class as plain-text injection.
Why it matters

A base64 string of 128+ characters appears in a documentation file. Encoded prompt injection hides the hostile instruction in base64 — invisible to keyword filters — and relies on the agent's ability to decode it at runtime. There is no normal authoring reason to embed a multi-hundred-byte base64 blob in skill docs.

The exact value spotted
excerptweb/public/data/authors/enessari.json· json
21".claude-plugin/marketplace.json": "{\n \"name\": \"metabase-ai-assistant\",\n \"owner\":
… (100 chars elided on L21)
22".claude-plugin/plugin.json": "{\n \"name\": \"metabase-ai-assistant\",\n \"description\":
… (100 chars elided on L22)
23"README.md": "<div align=\"center\">\n\n# 🚀 Metabase AI Assistant\n\n### **The Most Powerfu
… (101 chars elided on L23)
24},
25"plugins": [
Occurrences
1 occurrence · at L23
How to fix
Remove the encoded blob, or decode it and review what it actually contains.
  1. Decode the base64 string and confirm it is not an instruction directed at the agent.
  2. Move any legitimate binary or signature data into a dedicated file (*.sig, SIGNATURES) outside the documentation.
Framework references
OWASPLLM01ATLASAML.T0051
Trace & refs
ruleSS-SKILL-INJECT-B64-PAYLOAD-01sha2569d716118214b0ad4rubric 365aacaView on GitHub
HIGH"Ignore previous instructions" command embedded in the skillSS-SKILL-INJECT-IGNORE-01 · Prompt injection · web/public/data/authors/earthmanweb.json×3
HIGHwhen it fires on hostile content the impact is full system-prompt override.
Why it matters

The text {match} is the classic direct prompt-injection phrasing. Placed in a skill body that the agent reads as trusted instructions, it tries to make the agent abandon its prior rules and follow whatever comes next — a full system-prompt override.

The exact value spotted
excerptweb/public/data/authors/earthmanweb.json· json
66".claude-plugin/marketplace.json": "{\n \"name\": \"claude-flow-marketplace\",\n \"owner\"
… (100 chars elided on L66)
67".claude-plugin/plugin.json": "{\n \"name\": \"claude-flow\",\n \"version\": \"2.5.0\",\n
… (100 chars elided on L67)
68"README.md": "# 🌊 Claude-Flow v3: Enterprise AI Orchestration Platform\n\n<div align=\"cent
… (101 chars elided on L68)
69},
70"plugins": [
Occurrences
3 occurrences · first at L68, also L68, L68
Show all 3 locations
Line
File
L68
web/public/data/authors/earthmanweb.json
L68
web/public/data/authors/earthmanweb.json
L68
web/public/data/authors/earthmanweb.json
How to fix
Remove the override phrase, or rephrase the passage so it does not instruct the agent to discard its rules.
  1. Delete the ignore/disregard/forget … previous instructions sentence.
  2. If this is jailbreak-research or tutorial content, move the example into a clearly fenced, non-instruction block and label it as a quoted sample.
Framework references
OWASPLLM01ATLASAML.T0051
Trace & refs
ruleSS-SKILL-INJECT-IGNORE-01sha256087b391ca4342386rubric 365aacaView on GitHub
MEDIUMRequest for the agent to reveal its system promptSS-SKILL-INJECT-SYS-LEAK-01 · Prompt injection · web/public/data/authors/backbay-labs.json
MEDIUMit scores the disclosure intent; whether extraction succeeds depends on the host model's own mitigations.
Why it matters

The text {match} asks the agent to disclose its hidden system prompt or initial instructions. That is often the first step of a larger attack: knowing the system prompt lets an attacker craft inputs that defeat its constraints by mimicking its own voice.

The exact value spotted
excerptweb/public/data/authors/backbay-labs.json· json
20"files": {
21".claude-plugin/marketplace.json": "{\n \"name\": \"clawdstrike\",\n \"owner\": {\n \"n
… (100 chars elided on L21)
22"README.md": "<p align=\"center\">\n <img src=\".github/assets/clawdstrike-hero.png\" alt=\
… (100 chars elided on L22)
23"clawdstrike-plugin/README.md": "# ClawdStrike Claude Code Plugin\n\nRuntime security enforc
… (99 chars elided on L23)
24},
Occurrences
1 occurrence · at L22
How to fix
Remove the solicitation asking the agent to reveal its system prompt or hidden instructions.
  1. Delete the repeat/reveal/print your system prompt request from the skill.
  2. If you are debugging your own prompt, do so in a private dev harness rather than baking the request into a shipped skill.
Framework references
OWASPLLM07ATLASAML.T0051
Trace & refs
ruleSS-SKILL-INJECT-SYS-LEAK-01sha256d1f5bc1043a1dd23rubric 365aacaView on GitHub
Supply chainscore 100 · 0 findings
All supply chain checks passedNo findings in this category for the latest scan.pass
Maintenancescore 100 · 0 findings
All maintenance checks passedNo findings in this category for the latest scan.pass
Transparencyscore 100 · 0 findings
All transparency checks passedNo findings in this category for the latest scan.pass
Communityscore 100 · 0 findings
All community checks passedNo findings in this category for the latest scan.pass
Vendor response · right of reply
Are you the maintainer? Submit a response →

Audit the pieces. Scan the whole. Decide.

~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.