Security audit
Security checks for vulnerabilities and agentic risk
ClawScan rated this skill benign with medium confidence: it found no executable scripts or persistence indicators, while noting a mismatch between the stated purpose and a search capability, plus install and external-service risks.
The verdict is benign, not risk-free. Before installing, consider whether you want a skill that can search for listings as well as prepare and submit them, and whether you are comfortable with its unpinned tool references and varg API use. The scan was partial: it covered all four listed components, but one local reference could not be resolved, so the evidence is incomplete.
The supplied command calls the /v2/skills/search endpoint with a query and result limit. It does not inspect a skill folder, validate standards, assemble a listing, or submit one for review. Although it uses the same service, its observable purpose is skill discovery, explicitly outside the declared scope.
Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.
## 3. Check it against the standard
Run the checklist in `references/standard-checklist.md`. Report what fails in plain words. Blocking: secrets in files,
`curl | bash` or install hooks, scripts calling unknown hosts, bundled media without a license, no SKILL.md description.
Everything else (evals, metadata, ≤ 500 lines) is advice: list it, offer to fix it, publish anyway if the person wants.
## 4. Fill the card
Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.
- [ ] `SKILL.md` exists with frontmatter `name` and `description`.
- [ ] No secrets: no API keys, tokens or `.env` files in the folder or its git history you can see.
- [ ] Scripts are safe: no `curl | bash`, no install or post-install hooks, network calls only to hosts the skill names
(model APIs, its own CDN). Read every script.
- [ ] Bundled media has a license (own work, CC, or the vendor's terms). No film stills, no scraped photos of people.
- [ ] The source is public over https.
npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
homepage: https://varg.ai/how-to/publish
compatibility: >-
Publishes through the varg MCP server (https://mcp.varg.ai/mcp, tools publish_skill and
my_skills) or, without MCP, POST https://api.varg.ai/v2/skills/submissions with VARG_API_KEY
(or ~/.varg/credentials). Needs a varg account; publishing is free.
allowed-tools: Bash(curl:*) Read
---
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
# The varg how to card
Every listing on varg how to has this card. Fill every field you can verify; leave out what you cannot. These are the
fields of the varg MCP `publish_skill` tool and of `POST https://api.varg.ai/v2/skills/submissions`.
## Fields
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
Search with the skill's purpose in one line, through search_skills or:
curl -s "https://api.varg.ai/v2/skills/search?q=<what it makes, URL-encoded>&kind=skill&limit=5"
The same skill already listed (same repo, or the same thing under another name) → tell the person and do not publish a
npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
KEY="${VARG_API_KEY:-$(grep -o '"api_key"[^,}]*' ~/.varg/credentials 2>/dev/null | head -1 | cut -d'"' -f4)}"
curl -s -X POST https://api.varg.ai/v2/skills/submissions \
-H "Authorization: Bearer $KEY" -H 'content-type: application/json' \
-d '{"name":"…","summary":"…","source_url":"https://…","install":"…","use_when":["…"],"category":"…","tags":["…"],"via":"api"}'
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
KEY="${VARG_API_KEY:-$(grep -o '"api_key"[^,}]*' ~/.varg/credentials 2>/dev/null | head -1 | cut -d'"' -f4)}"
curl -s -X POST https://api.varg.ai/v2/skills/submissions \
-H "Authorization: Bearer $KEY" -H 'content-type: application/json' \
-d '{"name":"…","summary":"…","source_url":"https://…","install":"…","use_when":["…"],"category":"…","tags":["…"],"via":"api"}'
npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.
npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.