ctf-osint is the open source intelligence skill of the ctf-skills collection. It gives an agent an index of techniques for CTF challenges that ask it to find information in public sources, with each topic pointing to a supporting file.
The skill covers DNS and WHOIS lookups, image metadata, reverse image search and geolocation, social media and username research, Google dorking, web archives and a few platform-specific lookups. It lists optional Playwright MCP use for pages that need a browser.
What You Get
- A prerequisites list for Python packages and command line tools such as whois, dig, nmap and exiftool
- Quick-start commands for DNS, image metadata, web archive and Shodan lookups
- Short technique entries that link to four detailed reference files
- Pivot rules to the forensics, web and malware skills
- A list of public resources such as Shodan, Censys, VirusTotal, WHOIS and the Wayback Machine
Common Use Cases
- Working out where a photo was taken from signs, landmarks and map data
- Tracing a username or a social media account across platforms in a challenge
- Reading DNS TXT records and archived pages of a challenge domain for hidden clues
- Decoding coordinate formats such as MGRS and Plus Codes
What Makes It Unique
The skill is built around the question types that competitions ask: place a photo, follow a handle, read a domain’s records, identify an unknown string. It records where known shortcuts fail, for example sites that only work in a browser and platform false positives, and it points to browser-driven research through Playwright MCP.
Under The Hood
The skill is instructions only, with no scripts of its own. It tells the agent to run lookups with dig, whois, exiftool, nmap, curl and the Shodan command line, to call public APIs and fetch web pages and archives, and to use a browser through Playwright MCP when a page needs it. Its frontmatter allows shell, file writes, web fetch and web search. The reference files describe research on public accounts and records within a competition.