viktor-milev

business-decoder

Build a first-principles, plain-English understanding of any publicly listed company — the front-of-funnel lens applied BEFORE any valuation or thesis work. Covers what they do, how they make money, the need they serve, their customers, value-chain seat, ownership and control, the tailwind or headwind they ride, and whether it is a good, mediocre, or fragile business — translating ALL domain jargon (biology, physics, chemistry, engineering) into plain analogies a finance generalist gets. Trigger on "decode this business/company," "what does X actually do," "how does X make money," "understand this business," "is X a good business," "explain this company in plain English," "first-principles read on X," "I just heard about X." Industry-agnostic — biotech, photonics, industrials, software, anything. NOT for valuation, multiples, margins, or estimates (use a dedicated valuation skill); NOT for a buy/sell/short/mispricing view or sizing (use investment-council); NOT for fund or manager due diligence.

fable-gpt

Cross-model coding workflow for a user who cannot read code. Use Codex (GPT-5.6-sol) as an ADVERSARIAL AUDITOR of code written by Fable/Opus — and later as an executor for heavy implementation. Invoke at the start of any coding session that should get an independent second-model review. Triggers include "fable-gpt", "audit this with Codex", "adversarial review", "have Codex check this", "cross-model review". Do NOT trigger outside Claude Code (requires the codex-plugin-cc plugin and an authenticated Codex CLI), for non-coding requests, or before any implementation exists to audit.

investment-council

Pressure-test an investment thesis, trade, position, or mispricing claim through a council of five orthogonal advisors — Bear, Bull, Base Rate Skeptic, Market Structure Analyst, Risk Manager — who analyze independently, peer-review anonymously, and return a verdict tuned to DECISION (take/size this position) vs ANALYSIS (is this asset mispriced). Trigger on "council this trade/position/thesis/investment," "investment council," "pressure-test/war-room/stress-test this position," and on capital-at-risk asks like "should I buy/sell/short X," "is X mispriced/cheap/a good trade," "what's wrong with this thesis," "sanity-check this trade," "help me size this." For single securities, funds, portfolios, and mispricing claims. NOT for macro commentary without a position, portfolio construction, or fund-manager due diligence.

prompt-architect-fable

Fable-variant prompt architect — classifies a task against a 10-bucket taxonomy with an execution-mode overlay (interactive vs. agentic), decides output architecture (single-turn vs. chained multi-stage), and produces a technique-grounded v1 draft prompt calibrated for Claude Fable 5, with surface-aware run settings (chat app, Claude Code, Cowork, harness). Verdicts are SINGLE-TURN, SINGLE-TURN (CEILING), or CHAIN (README + Project orchestrator + per-stage prompts). Fires ONLY on explicit Fable invocation — trigger phrases include "architect this for Fable 5", "help me prompt this for Fable", "draft a Fable 5 prompt for". Do NOT trigger on unqualified requests ("help me prompt this", "draft a prompt for") — those route to prompt-architect (Opus sibling). Do NOT trigger when a draft prompt already exists — that's prompt-optimizer-fable. Do NOT trigger for tasks covered by a dedicated domain skill — route to that skill instead.

prompt-architect

Classifies a task against a 10-bucket taxonomy, decides output architecture (single-turn vs. chained multi-stage), and produces a technique-grounded v1 draft prompt. Verdicts are SINGLE-TURN, SINGLE-TURN (CEILING) with a strain note, or CHAIN producing a README + Project orchestrator + per-stage prompts. Use when you describe a non-trivial task but have no prompt written yet. Trigger phrases include "help me prompt this", "set up a prompt for", "draft a prompt for", "how should I ask Claude to", "architect a prompt for", "I'm about to start work on X — help me frame it". Also trigger for multi-step analytical tasks (research, knowledge document, design, decision support) where no draft exists. Do NOT trigger when a draft prompt already exists — that's prompt-optimizer. Do NOT trigger for tasks already covered by a dedicated domain-specific skill — route to that skill instead.

prompt-optimizer-fable

Fable-variant rigorous prompt optimizer — audits a polished prompt and rewrites it to elicit state-of-the-art output from Claude Fable 5, with surface-aware run settings (chat app, Claude Code, Cowork, harness). Triages into OUT OF PURVIEW / BORDERLINE / IN PURVIEW; decides output architecture (single-turn vs. chained multi-stage) and produces a chain deliverable (README + orchestrator + stage prompts) when signals fire. Fires ONLY on explicit Fable invocation — trigger phrases include "PO this for Fable", "optimize this prompt for Fable 5", "fable-optimize this", "run the fable optimizer". Do NOT trigger on unqualified requests ("PO this", "optimize this prompt") — those route to prompt-optimizer (Opus sibling). Do NOT trigger for messy dictated or flow-state input (use prompt-optimizer-flow), and do NOT trigger when the user wants the prompt EXECUTED rather than optimized.

prompt-optimizer-flow

Flow mode prompt optimizer — transforms rough, dictated, or stream-of-consciousness thoughts into a state-of-the-art Claude keystone prompt in seconds, without breaking the user's flow. Produces a finished, paste-ready prompt with no placeholders or template variables. Skips diagnostic theater, infers intent aggressively from messy input. Use whenever the user pastes or dictates messy unstructured input and wants it turned into a polished prompt fast. Trigger phrases include "quick prompt", "dictating", "rough thoughts", "thinking out loud", "PO this fast", "flow mode", "PO flow", "make this a prompt", "turn this into a prompt", "draft a prompt for", "starting a new chat about". Also trigger on dictation artifacts (rambling, self-corrections, "um", "uh", "actually wait") or when starting a new chat and needing a starter prompt. For polished prompts needing rigorous evaluation, use prompt-optimizer instead. Do NOT trigger when the user wants the prompt EXECUTED.

prompt-optimizer

Rigorous mode prompt optimizer — audits a polished prompt, decides output architecture (single-turn vs. chained multi-stage), and rewrites it to elicit state-of-the-art output from Claude. Triages every input into OUT OF PURVIEW / BORDERLINE / IN PURVIEW; rigorous optimization (architecture decision, 9-dim diagnostic, archetype detection, capability activation) applies ONLY to IN PURVIEW. When chain signals fire, produces a chain deliverable (README + Project orchestrator + per-stage prompts) instead of a single-turn prompt. Simple prompts return substantially unchanged with a triage note. Use when the user asks to optimize, audit, score, diagnose, sharpen, or level up a prompt; trigger phrases include "optimize this prompt", "PO this", "run this through the optimizer", "sharpen this prompt". For messy dictated or flow-state input, a separate flow-mode counterpart exists. Do NOT trigger when the user wants the prompt EXECUTED rather than optimized.

code-auditor-cc-codex