Pesquisa e Web

MCP client: connect servers, register tools (stdio/HTTP).

Discovers and invokes agent skills. Use when starting a session or when you need to discover which skill applies to the current task. This is the meta-skill that governs how all other skills are discovered and invoked.

Turns research notes, interview transcripts, or article text into a broadcast-ready voice-over script — short sentences, active voice, written for the ear, timed to picture.

Produces a structured research framework mapping the internal hierarchy, key decision-makers, departments, and reporting lines of any company, institution, or organisation — so journalists can understand who holds power and how decisions get made before conducting interviews or publishing a story.

Triage issues through a state machine driven by triage roles. Use when user wants to create an issue, triage issues, review incoming bugs or feature requests, prepare issues for an AFK agent, or manage issue workflow.

Evaluates the methodology of a published study in plain language — identifying its strengths, weaknesses, and the specific claims the study design can and cannot support — so a journalist can report on the research without overstating or understating its significance.

OBLITERATUS: abliterate LLM refusals (diff-in-means).

Writes a complete YouTube video description with the first two lines optimized for the search preview, followed by a structured body that serves both the algorithm and the viewer.

Produces a structured verification checklist for any scientific paper or research claim you plan to cite, covering every step a journalist should take to confirm the paper is valid, current, and not disputed or retracted.

Generates a structured biographical research plan for a specific person, listing the public-record sources to check, the biographical questions to answer, and the verification steps to confirm key claims — so the journalist walks into an interview or writes a profile with no blind spots.

Writes a clear, technical brief describing exactly what data needs to be collected from a website or set of web pages, how it should be structured, and what edge cases and legal/ethical considerations apply — for handoff to a developer or data team.

vLLM: high-throughput LLM serving, OpenAI API, quantization.