Search the live UK software/tech jobs board.

Args: query: free-text over title / company / location / employer. city: one of the 14 supported cities (e.g. "London") or "other" for remote/UK-wide. min_salary: minimum annualised GBP (matches stated salary only). sponsor: only jobs that explicitly offer visa sponsorship. licensed: only employers holding a UK Skilled Worker sponsor licence. remote: only fully-remote roles. discipline: a job-taxonomy discipline (see list_filters), e.g. "Backend". role: "engineer" or "manager". tech: comma-separated technologies (overlap match), e.g. "Python,AWS". level: comma-separated seniorities: intern,junior,mid,senior,staff,staff_plus. source: comma-separated source keys (see list_filters), e.g. "ats,linkedin". sort: "recent" (default), "pay", or "relevance". page: 1-based page. page_size: results per page (max 100; default 10 to keep output small). include_description: include the full job description for each result (large).

Returns a page of jobs with salary, sponsorship, tech, city, apply_url and detail_url. Use get_job(source, id) for a single full record.

Fetch one job by its source + id (e.g. from a search result's source and id), with the full job description. Returns {error} if it's not currently listed.

Look up a company by its company_key (from a job's company_key). Returns crawled facts (rating, size band, industry, logo, Companies House), Levels.fyi comp (USD/GBP), and an aggregate of its CURRENT live roles (count, top tech/disciplines, cities, a sample).

Match a CV against the live board and return the best-fitting roles (semantic retrieve → rerank → comp-weighted ranking). Stateless and anonymous — nothing is stored.

Args: cv_text: the CV as plain text / markdown (≥120 chars). city / min_salary / sponsor: optional filters on the candidate side. sort: "balanced" (default), "relevance", or "pay". top: max matches to return.

Returns {profile, matches}: a parsed board-aligned profile (disciplines, technologies, seniority) and ranked matches each with a relevance score and shared disciplines/tech. Rate-limited (~12/hr per IP).

ATS-tailor a CV to ONE specific job and return applyable edits (before → after diffs).

Give the job either as free text (job_text) OR as a board listing (source + job_id from a search result). Provide cv_text as plain text / markdown.

Returns: advice: labelled guidance (keyword MATCH, ADD THESE KEYWORDS, TITLE, FORMAT CHECK). edits: a list of {title, rationale, before, after} — each before is a verbatim CV snippet and after its ATS-tailored replacement; apply them as diffs. examples: anonymised real-world phrasing references (style only — never copy facts).

Never invents experience. Anonymous = up to 5 edits (free tier). Rate-limited (~30/hr per IP).

List the accepted filter values for search_jobs (cities, disciplines, levels, sources, common technologies, roles). Mirrors the board's canonical options.