Giovanni: Agent for Claude Code

.claude/agents/prediction-architect.md

prediction-architect is an agent for Claude Code from jaroslavsoucek-art/Giovanni. It costs 92 tokens per session (3,769 once invoked), scanned A, original, MIT.

A specialist for designing prediction features in an AI chief-of-staff system. It covers simulations of how stakeholders may respond, hidden testable predictions, and later accuracy reviews.

In plain words
What is it for?
It is for building stakeholder simulations, quarterly reviews of hidden predictions, and monthly accuracy scoring for each person.
Why use it?
It provides a structured way to explore possible reactions and measure whether predictions about people were accurate over time.

Agent for Claude Code

Written for Claude Code: installed under .claude/. Also seen: model in frontmatter; reads .claude/ paths; mentions CLAUDE.md.

This is jaroslavsoucek-art/Giovanni's own configuration. It tells Claude Code how to work on Giovanni itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything Giovanni configures →

Not installable: its command points at a path on the author’s own machine, so it runs nowhere else. The line is /Users/soucek/dev/Giovanni/memory/templates/stakeholder.template.md.

Reuse

Borrowing it

Nothing to install: this file belongs to jaroslavsoucek-art/Giovanni. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/jaroslavsoucek-art/Giovanni/main/.claude/agents/prediction-architect.md
Clone the repo
git clone --depth 1 https://github.com/jaroslavsoucek-art/Giovanni

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for prediction-architect

README.md
[![agentmods](https://agentmods.dev/badge/agents/jaroslavsoucek-art/giovanni/prediction-architect/github.svg)](https://agentmods.dev/agents/jaroslavsoucek-art/giovanni/prediction-architect)
Your own site
<a href="https://agentmods.dev/agents/jaroslavsoucek-art/giovanni/prediction-architect"><img src="https://agentmods.dev/badge/agents/jaroslavsoucek-art/giovanni/prediction-architect/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for prediction-architect

Your own site · 80×15
<a href="https://agentmods.dev/agents/jaroslavsoucek-art/giovanni/prediction-architect"><img src="https://agentmods.dev/badge/agents/jaroslavsoucek-art/giovanni/prediction-architect.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 92 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,769 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00092 $0.03769
Opus 5 $0.00046 $0.01885
Sonnet 5 $0.00018 $0.00754
Haiku 4.5 $0.00009 $0.00377

Measured today against content hash 60c88d732160, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

prediction-architect scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/agents/prediction-architect.md · 209 lines

How it starts

The opening of the file, as written. The whole thing — 209 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Prediction Architect (Giovanni specialist)

You own the predictive layer of Giovanni. This is the strongest IP moat in the entire framework — per gap analysis vs all known competitors and platform vendors, no one ships per-stakeholder predictive simulation with 3-tier no-percentages framing, anti-self-fulfilling shadow hypotheses, or actor-level calibration scoring. Your output is what makes Giovanni structurally different from alfred_, Lindy, Bond, Murchison's template, or anything Google/Anthropic will ship in the next 24 months.

Get this right.

Source

Live reference implementation — resolve the path with bash scripts/source-path.sh (reads $GIOVANNI_SOURCE, else the gitignored .source-path). Read-only. Never write there, never commit there.

It is a running system, not a copy. A write to it is damage to somebody's live operational state, not a dirty working tree — and unlike a snapshot, nobody will notice it as "the extraction scratch dir".

Two consequences of reading live rather than from a frozen copy:

  • Record what you read. Put the source HEAD (bash scripts/source-path.sh --sha) in whatever you produce. The source moves; a pattern extracted from an unnamed commit cannot be re-checked later, and "the source does X" with no commit behind it rots exactly like any other unanchored claim.
  • The no-leak rule matters more, not less. A live operational repo is full of current names, numbers and decisions. Everything you carry across is structure; anything that identifies the domain — person, company, product, market, vendor, codename — becomes a placeholder before it lands in this repo.

Key sources:

Predictive layer state:

  • memory/branch-out/ — simulation artifacts (read all 5 files). Note governance section draft.
  • memory/shadow/{pending,resolved,expired}/ — invisible hypotheses (read structure, sample 2-3 from each subdir)
  • memory/calibration/actor-scores.yaml — per-actor accuracy state
  • memory/calibration/monthly/ — monthly report outputs
  • memory/triage-heuristic.yaml — triage rules

Read the full file on GitHub · 209 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. today Changed · +7 lines · +1 tokens per session 60c88d732160
  2. 12d ago First seen · 202 lines · 91 tokens per session scan A f6e743da2199

Subscribe to this mod's changes

prediction-architect is an agent published in the GitHub repository jaroslavsoucek-art/Giovanni (2 stars, last pushed today), licensed MIT. It adds 92 tokens to every session and 3,769 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other agents, from other repositories