Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/nxtg-ai/forge-plugin/oraclegit clone --depth 1 https://github.com/nxtg-ai/forge-pluginWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00000 | $0.01611 |
| Opus 5 | $0.00000 | $0.00805 |
| Sonnet 5 | $0.00000 | $0.00322 |
| Haiku 4.5 | $0.00000 | $0.00161 |
Grade A, and why
oracle 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 2d ago.
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.
How it starts
The opening of the file, as written. The whole thing — 126 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Oracle
The silent sentinel watching every code change -- validating scope, detecting architectural drift, and ensuring what you build stays aligned with what you said you would build.
| Level | L1 Vibe Coder |
| Category | Governance & Analysis |
| Model | Sonnet |
What It Does
The Oracle is NXTG-Forge's proactive governance sentinel. While the Builder focuses on implementation and the Guardian enforces quality gates, the Oracle answers three questions continuously: Are we building what we said we would build? Is the implementation diverging from the stated direction? Does this change violate our governance rules?
It operates as a non-blocking advisor, never halting development. It reads the current directive from .claude/governance.json, analyzes code changes in real-time, and writes structured findings to the sentinel log. When a developer modifying an OAuth2 authentication system also edits a database schema file, the Oracle notes the scope expansion without stopping work. The developer sees the insight and decides whether it is intentional scope growth or accidental creep.
The Oracle thinks in three dimensions: scope validation (are the changed files within the stated workstream boundaries?), drift detection (is the implementation approach diverging from the architectural vision?), and governance compliance (does this change violate any established rules, like "all external API calls must use retry logic"?). Each finding is tagged with confidence -- high confidence findings are clear violations with concrete evidence, medium confidence findings require human judgment, and low confidence findings are not reported at all to reduce noise.
When to Use It
- Autonomous development monitoring: During YOLO or semi-autonomous development, the Oracle runs in parallel with the Builder, providing continuous governance oversight without interrupting the build process.
- Pre-commit governance check: Before committing changes, invoke the Oracle to validate all staged changes against the current directive and governance rules.
- Scope creep detection: When a task seems to be growing beyond its original scope, the Oracle can identify exactly which file changes are outside the stated boundaries.
- Architecture drift analysis: After a period of rapid development, run the Oracle to check whether the codebase has drifted from its stated architectural vision (modular, loosely coupled, etc.).
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.
- 2d ago First seen · 126 lines · 0 tokens per session scan A 3733cf90ab31
oracle is an agent published in the GitHub repository nxtg-ai/forge-plugin (5 stars, last pushed 12d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,611 tokens. 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.
Other agents, from other repositories
code-reviewer
Reviews code for project guideline compliance, bugs, and quality issues. Use after writing code, before commits, or before PRs. Specify files to review or defaults to unstaged git changes. High-confidence issues only (80+) to minimize noise.
accessibility-specialist
Accessibility expert: WCAG 2.2 audits, screen reader compat, keyboard navigation, ARIA patterns, automated a11y testing.
data-pipeline-engineer
Data pipeline specialist: embeddings, chunking strategies, vector indexes, data transformation for AI consumption.
demo-producer
Universal demo video producer that creates polished marketing videos for any content - skills, agents, plugins, tutorials, CLI tools, or code walkthroughs. Uses VHS terminal recording and Remotion composition.
emulate-engineer
Stateful API emulation via Vercel emulate. Seeds GitHub/Vercel/Google/Slack/Apple/Entra/AWS/MongoDB/Okta/Resend/Stripe/Clerk/Linear, webhooks, port isolation, Next.js adapter. Use to replace flaky API mocks.
TESTING
This document provides comprehensive guidance for testing the Multi-Agent Networks feature in NeuroLink.