tweet-thread-from-blog — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited tweet-thread-from-blog (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Convert a blog post or article into a 7-10 tweet thread. One insight per tweet. Strong hook. CTA in the final tweet. Optionally post the thread directly to X via Composio.
Apply these rules to every tweet you write:
Banned words — do not use any of these: incredible, amazing, leveraging, synergize, game-changing, game changer, let's dive in, buckle up, it's worth noting, delve, harness, unlock, groundbreaking, cutting-edge, remarkable, paradigm, revolutionize, disruptive, transformative, thrilled, excited to share, powerful, innovative, comprehensive, actionable, crucial, vital, pivotal, elucidate, utilize, can, may, just, that, very, really, literally, actually, certainly, probably, basically, could, maybe
If a draft tweet contains any banned word, rewrite it before presenting.
Do not invent specifics. Every claim, stat, and example in the thread must come from the blog post. Never fabricate data, quotes, or outcomes.
Confirm input is present. The user must provide one of:
If no input is present, ask: "What blog post or article should I turn into a thread? Share a URL or paste the content."
COMPOSIO_API_KEY is only needed for direct posting to X. Output-only mode works without it.
If input is a URL: Use WebFetch or Chrome DevTools MCP to fetch the page. Extract:
If input is pasted text: Read it directly.
After fetching, identify:
QA checkpoint: State the core thesis and list the top insights you will use. Confirm every insight comes directly from the source. Do not proceed if you cannot verify a claim.
Four styles. Auto-detect from content signals. Respect explicit user override.
| Style | When to Use | Signals |
|---|---|---|
| Data/Insight | Evidence-based article with stats or research findings | Numbers, percentages, study results, data points |
| How-To | Tutorial, guide, or step-by-step article | Numbered lists, step headers, "how to" in title |
| Story/Journey | Personal experience, build log, lessons learned | First-person narrative, "I learned", "we built" |
| Hot Take | Opinion piece, contrarian argument | "Why X is wrong", "Stop doing X", counterintuitive claim |
Decision logic:
State the chosen style and reason. If ambiguous, pick one and note it.
Read references/thread-format.md in full before writing any tweet. Internalize:
Read references/output-template.md and select the template matching the chosen style.
Produce three things:
(A) The full thread (7-10 tweets)
Each tweet:
(B) The CTA tweet (final tweet)
(C) One alternative hook tweet
Run every check before presenting. Fix violations first.
State total tweet count and confirm every tweet is under 280 characters before presenting.
Check for COMPOSIO_API_KEY.
If set: "Thread ready. Confirm to post to X via Composio, or say 'output only' to get the text."
On confirmation:
TWITTER_CREATION_OF_A_POST with text set to tweet 1 contentreply_in_reply_to_tweet_id = tweet 1 ID, capture its IDIf not set: Present each tweet in a numbered list inside a code block. Add:
"To enable direct posting, add COMPOSIO_API_KEY to your .env file. See README.md for Composio setup."
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.