Skip to content
Back to skills

Oracle

ASecurity

Oracle second-model review: bundle prompts/files for another AI, debug, refactor, design.

  • 10 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 10, 2026
ai-agentsbashnodegitapi

Works with

  • cli
  • api

Security analysis

A100/100

Scanned October 10, 2026

npx -y skills add mouadja02/skills --skill oracle --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Oracle?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Oracle
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/mouadja02-oracle/badge)](https://www.skillsdirectory.com/skills/mouadja02-oracle)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: oracle
description: "Oracle second-model review: bundle prompts/files for another AI, debug, refactor, design."
source: "https://github.com/steipete/agent-scripts"
attribution: "steipete/agent-scripts by Peter Steinberger"
version: "1.0.0"
---

# Oracle (CLI) — Second-Model Review

> **Attribution:** Sourced from [steipete/agent-scripts](https://github.com/steipete/agent-scripts) by [Peter Steinberger](https://github.com/steipete).

## When to Use

- Getting a second-model review of code, architecture, or design
- Bundling prompts and files into a single request for another AI model
- Cross-validating agent output with an independent model

Oracle bundles your prompt + selected files into one "one-shot" request so another model can answer with real repo context (API or browser automation). Treat outputs as advisory: verify against the codebase + tests.

## Main Use Case

Default workflow: `--engine browser` targeting GPT-5 Pro in ChatGPT. This is the "human in the loop" path.

## Commands

```bash
# Show help
npx -y @steipete/oracle --help

# Preview (no tokens spent)
npx -y @steipete/oracle --dry-run summary -p "<task>" --file "src/**" --file "!**/*.test.*"

# Token/cost sanity
npx -y @steipete/oracle --dry-run summary --files-report -p "<task>" --file "src/**"

# Browser run (main path; long-running is normal)
npx -y @steipete/oracle --engine browser --model gpt-5-pro -p "<task>" --file "src/**"

# Manual paste fallback (assemble bundle, copy to clipboard)
npx -y @steipete/oracle --render --copy -p "<task>" --file "src/**"
```

## Attaching Files (`--file`)

`--file` accepts files, directories, and globs. Pass it multiple times:

```bash
# Include directory glob
--file "src/**"

# Include specific file
--file src/index.ts

# Exclude (prefix with !)
--file "src/**" --file "!src/**/*.test.ts" --file "!**/*.snap"
```

Default-ignored dirs: `node_modules`, `dist`, `coverage`, `.git`, `.turbo`, `.next`, `build`, `tmp`.

## Budget + Observability

- Target: keep total input under ~196k tokens.
- Use `--files-report` and/or `--dry-run json` to spot the token hogs before spending.

## Sessions + Reattachment

Sessions stored under `~/.oracle/sessions`. Runs may detach or take a long time.

```bash
# Don't re-run; reattach
oracle status --hours 72
oracle session <id> --render
```

## Prompt Template (high signal)

Oracle starts with **zero** project knowledge. Always include:

- Project briefing (stack + build/test commands + platform constraints).
- "Where things live" (key directories, entrypoints, config files).
- Exact question + what you tried + the error text (verbatim).
- Constraints ("don't change X", "must keep public API", "perf budget").
- Desired output ("return patch plan + tests", "list risky assumptions", "give 3 options with tradeoffs").

## Safety

- Don't attach secrets by default (`.env`, key files, auth tokens).
- Prefer "just enough context": fewer files + better prompt beats whole-repo dumps.
- **API runs require explicit user consent** before starting because they incur usage costs.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…