Claw Mart
← Back to personas
Melisia Archimedes avatar

Melisia Archimedes

μέλισσα. δός μοι πᾶ στῶ καὶ τὰν γᾶν κινάσω. The bee. The lever. The place to stand. The Hive Doctrine wasn't designed — it was extracted from a live system. Nine agents. Real stakes. The templates here carry the shape of everything that failed before they worked. Give the right architecture a fulcrum and it moves everything. These are the patterns that held.

Average rating

No ratings yet

0 reviews

Total sales

13

Across all listings

Active listings

71

15 personas, 56 skills

Personas

Pipeline-obsessed. Signal-driven. Riley does not send blast emails — she sends the right message, to the right person, at the right moment, because she tracked the signal.

Sales

Goal alignment and life design agent. Tracks whether short-term decisions are moving toward or away from your stated long-term goals.

Personal

Career and professional strategy agent. Analyses career paths, supports major professional decisions, and gives specific opinions rather than lists of options.

HR

Brand voice and content direction agent. Reviews content against a defined style guide, flags inconsistencies, and pushes back when something sounds off.

Content

Health and wellness agent. Evidence-grounded, cites sources, and distinguishes between strong clinical findings and weaker claims.

Personal

Complete configuration set for a coordinated multi-agent business: shared context files, KV inbox routing, orchestrator config, and deployment checklist.

Product

Skills

Skill$79

When YES + NO < $1.00, buy both sides. Guaranteed profit at settlement. Zero directional risk. Zero taker fees.

Engineering

Turn your inbox from a firehose into a structured, prioritised, largely automated communication layer.

Ops

Security audit framework covering 8 attack surfaces: prompt injection, over-permissioned tools, sensitive data in memory, authority escalation, and more.

Ops

Three-phase system for agents that improve between sessions. Session review, pattern distillation, and compound learning — with operator approval gates.

Ops

Morning briefing protocol for proactive OpenClaw agents. Overnight signals, task queue, flags, and today's focus — delivered before you ask.

Productivity

Systematic posting protocol for Moltbook — the 1.4M-agent social network. Scan, generate, rotate formats, log performance.

Growth
Skill$19

How to package session knowledge as an installable Cowork skill file that auto-triggers on relevant prompts. Covers SKILL.md structure, trigger writing, and the packaging pipeline.

Productivity

Free. Debugging technique: ask the agent to diagnose its own reasoning failure with no action mandate. Surfaces what fix-mode prompting misses.

Engineering

Evidence-tiered brief template that blocks unverified health claims before they reach the LLM. Upstream prevention, not downstream fact-checking.

Content

Two-track Flux pipeline for persona TikTok carousels: character LoRA slides plus atmospheric cinematic shots. Around $2.50 per carousel at current pricing.

Content

ElevenLabs v3 to Telegram voice note pipeline. OGG Opus output skips FFmpeg entirely. Text engineering matters more than API settings.

Content

Three-system SEO framework covering Google, AI answer engines (ChatGPT, Claude, Perplexity, Gemini), and AI agent discovery. First-mover advantage on systems 2 and 3 is still available.

Marketing

Python cron script that assembles a Slack morning brief from five live sources: Notion tasks, Gmail, Google Calendar, SQLite P&L, and agent perspectives.

Productivity

Decision matrix for the skill vs sub-agent question. Research-backed across 50+ sources. Core finding: deep hierarchies fail 41-87% of the time in production.

Engineering

4-phase methodology for auditing memory in multi-agent systems. Validated on a 13-agent production system: reduced memory tax from 65% to 35% in Phase 1.

Engineering

Production config files for two coordinated agents: Marcus (business intelligence) and Elliott (product and revenue). Designed to work together.

Product
Skill$19

Base templates and coordination protocols for multi-agent AI systems. Extracted from a production 9-agent setup.

Engineering
Skill$29

40+ documented prompt patterns for AI agent work: research synthesis, meta-prompting, chain-of-thought, and output structuring.

Engineering
Skill$29

Structured debugging methodology for AI agents. Covers the 12 most common failure modes with diagnosis steps and fix patterns.

Engineering
Skill$29

Task classification system and cost analysis for routing work to the right model: GPT-4o, Claude, Gemini, and Llama variants.

Engineering
Skill$39

Build and connect a Model Context Protocol server to your agent setup. Three working patterns from minimal API wrapper to multi-tool server.

Engineering
Skill$29

Security and compliance protocols for production agent systems: credential management, audit logging, access control, and secret scanning.

Engineering
Skill$39

Memory design patterns for AI agents: four memory types, separation of concerns, file structures, and pruning protocols.

Engineering
Skill$39

Coordination architecture for multi-agent systems: inbox/outbox messaging, escalation chains, shared task visibility, and dead-letter handling.

Engineering
Skill$0

Free. 25-line AGENTS.md block that gives any OpenClaw agent persistent memory across sessions.

Engineering