All projects matching “persona”

Prompt-injection screen for text or a web page an agent is about to read: finds instruction overrides (many languages and obfuscated spellings), fake system or role markers, jailbreak personas, data-exfiltration links, requests for keys or seed phrases, tool, shell, or payment hijacks, and hidden content (invisible Unicode smuggling, zero-width and bidi characters, HTML comments, hidden elements,

For AI agents building agent swarms, synthetic users, NPCs, or trading bots. Input: archetype + prompt (e.g. 'risk-taking crypto founder'). Output: OCEAN Big Five + Dark Triad + trading psychology vector with emotional drift and memory anchors. M2M: use when you need consistent agent personality, backtest personas, or generate 1000s of synthetic traders. Pay-per-call USDC Base, no API key.

Search Paul Graham's source corpus — a source-grounded persona of the Y Combinator co-founder and essayist, built from 13,000+ cited sources: his complete essays, talks, interviews, and X posts — by semantic + lexical query and retrieve grounded packets (graded claims, quotes, evidence with citations). POST a JSON body { query, limit? (<=100), packetTypes? }. Flat micro-price per query.

Alex Hormozi advisor (Offers, pricing, and sales). Send POST with JSON body {"question": string (8+ chars), "context"?: string}; returns JSON {persona, answer, disclaimer, usage}. AI simulation based on publicly available information. This is not Alex Hormozi, is not written, reviewed or endorsed by Alex Hormozi, and does not speak for them. General information only, not financial, legal, tax or m

Paid MCP endpoint — POST a JSON-RPC tools/call. Tools: get_content / refresh_content (public profile data: github, twitter, linkedin) and the rep_* wallet-reputation tools (rep_rank, rep_rank_batch, rep_achievements, rep_persona_match, rep_profile_report). Flat USD prices per call, quoted in the 402.

Prompt-injection scan for text your AI agent is about to read (web pages, emails, tool output, MCP tool descriptions): risk score 0-100, verdict, categories and exact spans. Detects instruction overrides, jailbreak personas, system-prompt extraction, fake chat delimiters, data-exfiltration URLs, tool hijacking, hidden Unicode and encoded payloads. Deterministic rules, no data stored.