aaronguo/muse
Conversational writing topic discovery. Combines personal context (journals, notes, past writing), optional Signal DB intelligence, and web research to surface timely, authentic topic ideas for any writing format.
90
90%
Does it follow best practices?
Impact
—
Average score across 4 eval scenarios
High
Do not use without reviewing
Security
1 high severity finding. You should review these findings carefully before considering using this skill.
W007: Insecure credential handling detected in skill instructions.
What this means
The skill handles credentials insecurely by requiring the agent to include secret values verbatim in its generated output. This exposes credentials in the agent’s context and conversation history, creating a risk of data exfiltration.
Why it was flagged
The prompt tells the agent to read an auth token from EXTEND.md/.env or the environment and insert it verbatim into curl Authorization headers, which requires the LLM to handle and output secret values directly (high exfiltration risk).
Where we found it
[REDACTED]
secret · 2 sites
Step 2 instructs the agent to read the auth token from .env or environment variable and insert it verbatim into curl Authorization headers at two locations
SKILL.md
48
-H "Authorization: Bearer {token}" \
SKILL.md
57
-H "Authorization: Bearer {token}" \
Low
Low-risk findings.
1 low severity finding. Worth noting, but not necessarily harmful.
W011: Third-party content exposure detected (indirect prompt injection risk).
What this means
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
Why it was flagged
The skill's required workflow (SKILL.md Step 4 "Scan the Landscape" and Step 6d "WebFetch") explicitly performs targeted web searches and fetches top-ranking articles from public sites (X/Twitter, Reddit, Hacker News, YouTube, general web), and Step 2 can query a Signal DB that ingests public sources, so the agent reads untrusted, user-generated third‑party content which can change its subsequent recommendations.
Where we found it
x.com
domain · 2 sites
Step 4 explicitly searches X/Twitter for trending content; Signal DB also ingests from X/Twitter per setup docs
SKILL.md
108
| X/Twitter | `{pillar} trending {current_month} {year}` | The live pulse |
references/config/first-time-setup.md
120
> *"A Signal DB is a personal database that continuously collects news and discussions from sources like Hacker News, Reddit, X/Twitter, ArXiv, and GitHub Trend
news.ycombinator.com
domain · 2 sites
Step 4 explicitly searches Hacker News (news.ycombinator.com); Signal DB also ingests from Hacker News
SKILL.md
109
| Hacker News | `site:news.ycombinator.com {topic} {year}` | What builders care about |
references/config/first-time-setup.md
120
> *"A Signal DB is a personal database that continuously collects news and discussions from sources like Hacker News, Reddit, X/Twitter, ArXiv, and GitHub Trend
reddit.com
domain · 2 sites
Step 4 explicitly searches Reddit (reddit.com); Signal DB also ingests from Reddit
SKILL.md
110
| Reddit | `site:reddit.com {pillar} {current_month} {year}` | Where people argue |
references/config/first-time-setup.md
120
> *"A Signal DB is a personal database that continuously collects news and discussions from sources like Hacker News, Reddit, X/Twitter, ArXiv, and GitHub Trend
youtube.com
domain · 1 site
Step 4 explicitly searches YouTube (youtube.com) for video content
SKILL.md
112
| YouTube | `{topic} {current_month} {year} site:youtube.com` | Where the video gap is |
general web search
content-type · 1 site
Step 4 explicitly performs general web searches that fetch from untrusted public sources
SKILL.md
111
| General web | `{pillar} {year}` | The broader narrative |
WebFetch of top-ranking articles
content-type · 1 site
Step 6d explicitly uses WebFetch to fetch and read top-ranking existing articles on the same topic
SKILL.md
173
Once the angle is clear, use `WebFetch` on 1–2 top-ranking existing pieces on the same topic. Evaluate:
arxiv.org
domain · 1 site
Signal DB ingests content from ArXiv per setup documentation
references/config/first-time-setup.md
120
> *"A Signal DB is a personal database that continuously collects news and discussions from sources like Hacker News, Reddit, X/Twitter, ArXiv, and GitHub Trend
github.com
domain · 1 site
Signal DB ingests content from GitHub Trending per setup documentation
references/config/first-time-setup.md
120
> *"A Signal DB is a personal database that continuously collects news and discussions from sources like Hacker News, Reddit, X/Twitter, ArXiv, and GitHub Trend
- Audited
- Security analysis