mythos-proxy

Web Search

Search the web for current information. Executed server-side — returns a summarized answer with source URLs. Use for recent events, external docs, prices, news.

X Search

Search X (Twitter) posts, users and threads. Executed server-side — returns summarized results with post links. Use for social sentiment, announcements, public figures.

X User Search

Find X (Twitter) user profiles by name or handle. Executed server-side — returns matching accounts with handles, bios and follower counts. Use when looking for a person or organization.

X Keyword Search

Search X (Twitter) posts by exact keywords and operators (from:, since:, quotes). Executed server-side — best for precise queries like "from:handle" or specific phrases.

X Semantic Search

Search X (Twitter) posts by meaning rather than keywords. Executed server-side — best for natural-language questions, sentiment and topics that keywords miss.

X Thread Fetch

Fetch a full X (Twitter) thread/conversation by post ID. Executed server-side — returns the original post and its replies/quotes in order.

Web Fetch

Fetch a URL, executed server-side. Defaults to a plain GET; method/headers/body turn it into a full HTTP client for vendor-API exercises (e.g. POSTing to an operator-authorized API with the operator's own credentials). Returns the status line, content-type, and body (truncated) — error bodies included.

Image Gen

Generate an image from a text prompt (xAI Imagine). Executed server-side — returns the hosted image URL.

Intelx Search

Search Intelligence X (data-leak/darknet archive) for a selector: domain, email, IP, URL, BTC address. Discovery runs on the ANONYMOUS public tier — it never touches the paid 200/day search quota, and result listings include gated records (accesslevel 4/6) from PRO buckets (darknet.tor, leaks.logs, whois, dns…). Returns compact records with storageid/systemid for intelx_preview/intelx_read. Page deeper with intelx_results (free, unlimited).

Intelx Results

Page through an existing IntelX search by id — zero-cost, unlimited deep paging (replaces paid CSV export). Result ids are per-shard: pass shard='paid' only for searches created on the paid shard.

Intelx Preview

Quick look at a record's content (paid preview pool, 10000/day). Pass storageid/bucket/media/type from a search record. Gated content is unredacted when the paid key's entitlement covers the bucket.

Intelx Read

Read the full content of a record (paid read pool, 500/day — the chain's only scarce resource). Accepts storageid or systemid + bucket. storageids harvested from public-tier discovery work here — identifiers are global across shards.

Intelx Account

Show the paid IntelX key's entitlements (buckets) and remaining daily quotas per endpoint. Use before large harvests to budget the read pool (500/day).